mirror of
https://github.com/runbear-io/beardrive.git
synced 2026-08-25 08:08:08 +02:00
8931004b4a293d9822b097b95bc01846ea76ffe0
48
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
|
8931004b4a |
docs: use cases move to the marketing site, and the logo goes home (#153)
The seven use-case pages were the one part of docs.beardrive.ai that wasn't documentation — job-shaped routing pages, aimed at someone deciding rather than someone building. beardrive.ai already publishes six of them, same titles, same slugs, so the docs copy was a fork of marketing copy that no one was keeping in step. They're deleted here, and every URL redirects to its counterpart over there rather than to an index: those pages were live and indexed for a month, and sending all seven to /use-cases/ would throw away which one someone asked for. shared-skills is the exception — the landing site has no page for it yet, so that one lands on the index. The logo now points at beardrive.ai. Starlight always aims it at the docs root and has no config for it, so this is a SiteTitle component override: its own markup, one changed href, minus the two-logo branch this site never takes (one `src`, so the light/dark alternate can't render). Three off-site links — Use cases, Blog, GitHub — go at the BOTTOM of the sidebar. The sidebar order is the recommended path, so a link that leaves the docs must not sit above the docs. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
3d2acf658c |
ci(docs): docs.beardrive.ai deploys itself again (#152)
docs.beardrive.ai was serving a build from around 2026-07-19 — three weeks and 38 docs commits stale. /manual/hooks/ 404s while /manual/skills-and-hooks/, deleted in #85, still serves; the sitemap has no <lastmod> and robots.txt is Cloudflare's managed content-signals file with no Sitemap: line, so #140 plainly never shipped. The cause: the Pages project (beardrive-docs, docs.beardrive.ai) is a direct-upload project with no Git provider. Someone ran `wrangler pages deploy` by hand, then stopped, and nothing anywhere noticed — every check this repo has runs during a deploy that was no longer happening. docs.yml builds web/docs on PRs that touch it and deploys to Pages on pushes to main. Not the Pages Git integration, deliberately: it clones shallow, and astro.config.mjs reads each page's <lastmod> from the commit date behind it, so a depth-1 checkout drops all 27 of them — hence fetch-depth: 0. It also would rebuild the docs for every commit in a repo that is mostly Go. The path filter includes internal/webapp/frontend/src/tw.css: the palette is generated from that file, so it is a docs input even though it lives outside web/docs (which is also why the checkout can't be sparse). The post-deploy check:sitemap run is continue-on-error. Half of what it checks — a Cloudflare-managed robots.txt shadowing ours, cache propagation right after upload — isn't this repo's call, and a good deploy shouldn't go red over it. Needs CLOUDFLARE_API_TOKEN (Pages: Edit) and CLOUDFLARE_ACCOUNT_ID. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
37bd466bb7 |
docs: three claims that no longer match the code (#148)
- start/setup says "Three are shipped" above four templates. Four ship (internal/templates/files: docs, para, skills, wiki). - guides/shared-agent-memory tells the reader to run `/beardrive:install`. There is no such command — no plugin, no skill; the integration is internal/agenthooks and setup is the paste on start/setup. - index lists SSO as part of the managed service. No SAML/OIDC exists anywhere, and the cloud plan has it unstarted. Claude-Session: https://claude.ai/code/session_01UmhLKHq3QTHjkDi2wCDhNB Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
922886b949 |
fix(sync): a refused push stays refused, and says why (BEA-403) (#146)
A device whose journal push the hub 403s reported healthy sync between every pair of remote passes, and never showed the hub's reason for the refusal — so a user whose device was not registered to their account re-ran `bdrive login` (which the message told them to), re-checked their project permissions (write), and had nowhere left to look. Two causes, both local to the client: - Cycle recomputed st.Access from scratch at the end of every pass, including the daemon's cheap local-only ticks that never reach the hub. Three of those run between remote passes, so the daemon alternated "read-only on this project" / "access restored; syncing normally" every few seconds and `bdrive status` reported OK moments after a refused push. Now each leg records its own verdict — pull clears no-access, push records read-only or clears it — and a cycle that asked nothing leaves the last answer standing. - The hub's own sentence was summarized into "read-only (pull only)", which describes the STATUS CODE. It is the only thing that tells a device-registration refusal from a project the user really is a reader on. It now rides in SyncState.AccessReason and Result.Reason(), printed by `bdrive sync`, `bdrive status` and the daemon log, and dropped unless it passes journal.SafeText — hub text reaching a terminal. The hub's refusal also now names the upgrade: the binding is made by the login request naming its device, which a CLI older than the gate does not do, so "run `bdrive login`" alone sent that user in a circle. Hub and CLI deploy separately, so the skew is the expected state right after the gate ships. Claude-Session: https://claude.ai/code/session_01GSHsQU4pBCzKkPyPeXSwTm Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
5623113ff7 |
feat(sync): agents hear what teammates changed before they overwrite it (BEA-127) (#144)
A teammate's agent rewrites a file and yours never hears about it — the only trace is a `.bdrive-conflict-*` nobody opens. Now the turn-start hook names the paths that arrived since the last turn: "re-read before editing". The record lives in a spool (`internal/store/inbound.go`), not in Cycle's Result, because the daemon usually materializes a peer's change seconds before the turn starts — so the hook's own cycle sees nothing. materialize appends every path it writes or removes, `bdrive sync --hook` drains it after its cycle and renders it under each mount's own prefix (stripping the session subpath when the run is inside a mount, dropping paths outside it). Advisory only: nothing blocks, nothing prompts, no per-Write remote call. The spool is capped, 0600, in the volume dir, and best-effort everywhere — a spool failure never fails a cycle or a turn. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Snow Lee (Sungwon)andGitHub
|
2875e033be |
feat(webapp): mermaid fences render as diagrams, in the viewer and on share pages (BEA-91) (#143)
A ```mermaid fence rendered as a wall of `graph TD` source on both surfaces. It now renders as an SVG in the hub file viewer and on public /s/<token> markdown share pages. Mermaid ships inside the binary (no CDN, so air-gapped self-hosters keep working) and is imported lazily: a document with no fence downloads none of it. A fence that doesn't parse — the common case for hand-written wiki diagrams — keeps today's <pre><code> plus a small note, and one bad fence never stops the good ones beside it. A blocked or offline chunk lands in the same place. The share page is the harder half: it is server-rendered Go HTML with no JavaScript, and its `sandbox allow-scripts` CSP makes the origin opaque, so a module script and every import() it makes arrive with `Origin: null`. The CSP is unchanged and gains no allow-same-origin; instead the real-asset branch of frontend() now sets Access-Control-Allow-Origin, which only ever touches files that are already public and cookie-less. The script tag itself is injected only when the rendered document actually contains a fence. Also fixes an embed bug this change surfaced: `//go:embed static` silently skips names beginning with `_`, and Vite's first shared chunk is `_commonjsHelpers-<hash>.js`. The build passed, the commit looked right, and the served app was blank. It is `all:static` now, with a test that every file on disk is in the binary. |
||
|
|
4031495c81 |
feat(cli,docs): say that agent skills sync, and refuse ~/.claude as a mount root (BEA-117) (#138)
`.claude/skills/**` has always synced — deliberately, per the reservation rule's own comment — but the only sentence saying so sits under the heading "What beardrive does not sync". Nobody knows. Track B, the one real bug: `bdrive init ~/.claude` was accepted. The reserved-path rule matches ".claude/settings.json" on its directory segment, so at that mount root the file is bare "settings.json" — reserved by nothing — along with .credentials.json and every saved session under projects/. New exported config.AgentConfigDir folds the keys of agentHookConfigs the way ReservedDir folds (case, trailing dots), and init refuses before any network call or file write. Only that direction leaks: a mount CONTAINING ~/.claude still sees .claude/settings.json, reserved at any depth. Track A, the content job: a README Features bullet stating the positive claim, a 7th use-case page (plus its astro.config.mjs sidebar entry, without which it is invisible), and a `skills` template appended last to the registry so `docs` keeps the RECOMMENDED badge. The embed directive becomes `//go:embed all:files` — a plain pattern drops dot-prefixed paths silently, so the template whose whole payload is .claude/skills/<name>/SKILL.md would have shipped empty. templates_test.go's every-directory-holds-a-file rule now marks ancestors, not just the direct parent: skills is the first template more than one level deep, and the rule was stricter than its own stated reason (an intermediate directory on the way to a file is not empty). Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
d3d92bf904 |
feat(webapp): refuse to share a file that looks like it holds credentials (BEA-111) (#137)
Minting a share link ran zero content checks: a file holding an AWS-shaped key became a public URL on one click, and the CLI printed nothing but the link. handleShareCreate now reads the first 1 MiB and runs six anchored rules between the synced-path check and Shares.Create, answering 409 with rule ids and line numbers unless the request carries confirm: true. The matched text never leaves scanSecrets — not into the body, not into a log line. TestShareSecretNeverEchoed greps both for the planted string, because a 409 body is the easiest place in this codebase to leak it. Both callers carry the override, since the gate alone would turn any false positive into a hard block with no way out: `bdrive share --force`, and the browser's Share-anyway dialog on modalConfirm (no new component). A path that already has a live link skips the scan — its content is public already, so withholding the URL protects nobody — but alreadyPublic drops links whose creator left the org, since those 404 at /s/ and would otherwise wave a secrets file straight through. A failed blob read is 503, not a silent pass: the repo's "degrade rather than fail" posture is for sync cycles, and a check that skips itself on a storage hiccup is the false confidence this exists to remove. Every user-facing string says the file was checked at the moment you shared it. A link serves the file's LATEST content forever, so a key written into an already-shared file is never caught — that open loop stays open, and the copy is the only thing stopping v1 from claiming otherwise. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
594a027c15 |
feat(cli): bdrive grep — search the text inside the files a project syncs (BEA-99) (#136)
The ⌘K palette searches file names, projects and actions; nothing in the product searched file contents. Three personas independently typed a phrase that lives inside a synced file and got "No matches". `bdrive grep <pattern> [folder]` searches the working folder — RE2 or -F literal, -i, -l, -n (default 200, 0 = all), output `path:line: text`, exit 0 on match and 1 on none. It searches exactly what the project syncs, via a new syncer.SyncedFiles that wraps the existing walkFolder: the one copy of the sync predicate, so an ignore rule or a narrowed scope excludes a file from search the same way it excludes it from sync, and .bdrive/ state can never surface. Not Explain, which countFiles every pruned dir — a grep in a repo with node_modules/ would walk it in full for a count it discards. A read stays a read: LoadProject, not ResolveMount (no registry self-heal, no device enrollment), no session, no flock, and the volume store is opened for IgnoreAccepted only when it already exists, so a search creates nothing. Both the path and the matched line go through safeField — a matched line is a teammate's file content, the widest version of the surface that function exists for. The hub-side content index stays deliberately unbuilt; the issue records its cost. ROADMAP's "Search across the hub" line is reworded rather than removed. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Snow Lee (Sungwon)andGitHub
|
5f1ac98dae |
feat(hub): see what each agent session read, not just what it changed (BEA-98) (#135)
History showed what an agent run CHANGED. What it read lived in a daily aggregate with no session dimension, so the two could not be joined and nobody could answer "when my agent answered, what did it look at — and was it the fresh version or archive/retired-spec.md?". The join is one string carried through four places: hook -> spool -> hub -> run card. A run card now marks each change the run also read, lists the files it read and never touched, and says on screen why a read can be missing. The three landmines the issue asks be named here: 1. Op.Note is USER-SETTABLE (`bdrive sync --note`), so joining reads to writes on the note string would let any member with write access forge a note that collides with a teammate's run card and hang their reads off it. Fixed by adding journal.Op.Session — set only by `bdrive sync --hook`, never by --note — and joining on that. The note stays settable and stays untrusted; the join simply never reads it. Op.Session is additive JSONL and, like Mtime, is never an input to Less or Replay, so replay determinism is untouched and older ops carry "". The read half has the same hole one step further on: POST /reads takes the session id from the CLIENT, so a member could report reads under a teammate's session and paint files onto their card. Every session row is therefore pinned to the ownsDevice-validated device, and the query requires ?session= AND ?device= together — a forged row can only be found under the forger's own device, which MayActAs guarantees is never somebody else's. 2. BUCKET CARDINALITY. Putting the session in the read_stats key would take a 2k-file project from ~2k to ~100k rows/day, into a table ReadLedger loads whole at boot and full-scans on every heat request, hub-wide — so it would slow the Dashboard for projects that never ran an agent. This is the escape hatch the spec itself names, taken up front: session rows live in their own read_sessions repo, outside ReadLedger.byKey. No read_stats PK migration, no change to the resident-row count, ?by=device byte-identical. They get their own retention (session_retention_days, default 30) which DELETES rather than folds — no heat total was ever derived from them. 3. READS ARE RECORDED ONLY FOR PATHS IN THE CURRENT REPLAY, so a session that read a file it then deleted shows a change with no read. That is by design, and the run card says so in its footer rather than leaving it to read as a bug. Privacy ruling, written into internal/webapp/reads.go before anything serves it: a session id appears only in History responses on the op that carries it, and as a ?session= filter INPUT. It is never enumerated — no listing endpoint, no session column in /heat output, nothing new in ?by=device. Also: PendingReads now dedupes on (path, session), not path alone. Two agent sessions on one device between syncs used to collapse into one event carrying whichever session flushed last — one session's reads silently credited to another. Tests: journal round-trip + Less-ignores-Session; the forge test (`sync --note "claude-code session <someone-else's>"` leaves Session empty); a multi-device syncer test carrying the session through convergence; spool per-session dedup; hub round-trip, cross-device forge, query contract and non-enumeration; db_conformance on file, sqlite AND postgres; runs.ts grouping incl. legacy fallback; a Playwright spec on the seeded run card. |
||
|
|
e19fec0534 |
feat(webapp): old URLs follow a moved file, and share links follow the file (BEA-81) (#130)
There is no rename in beardrive: the scanner emits a put at the new path and a delete at the old, same device, same blob, one cycle. Everything keyed on a path therefore broke the moment a file moved — the viewer 404'd, history lost the file's own past versions, restore refused them, and a share link either 404'd or silently served whatever unrelated file later took its address. internal/webapp/moves.go derives the pairing from the ops the replay already walks, cached with the snapshot. Deliberately not a rename op: journal.Less and Replay are what every device converges to, and every already-shipped journal would still need the heuristic to read its own history. The two rules point in opposite directions on purpose. A viewer URL is an address, so a LIVE path always wins and only an empty one redirects. A share token is a promise about one file, so it follows the file even when a new one takes the old address — and 404s forever once the file is deleted. Nothing here writes an op or touches sync. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Snow Lee (Sungwon)andGitHub
|
7dd5d097fe |
docs: one answer on Windows support — macOS and Linux only (BEA-77) (#126)
"Can my Windows teammate join?" got three answers from three official surfaces: README's feature list said macOS & Linux, its CLI table documented an HKCU Run entry on Windows, and the hub's Installation page named no OS at all. Windows autostart is real code that cannot run — internal/store's flock and internal/daemon's Kill/Setsid are unix-only, so GOOS=windows does not build. The fix is the docs, not the port: every Windows claim comes out of README, web/docs and `bdrive autostart`'s own help, and the Installation page gains one line naming the supported systems. No Go source is deleted; internal/autostart/autostart_windows.go stays compiled-but-dormant for whenever the port happens. |
||
|
Snow Lee (Sungwon)andGitHub
|
70cf9818ce |
fix(docs): sitemap gains lastmod, and robots.txt points at it (#140)
docs.beardrive.ai shipped no robots.txt, so nothing on that host named the sitemap — a crawler arriving at the subdomain had to guess the URL or be handed it in Search Console. And every entry was a bare <loc>: no freshness signal at all, on the site whose whole value is being current. Each URL now carries the commit date of the markdown behind it, read from one `git log` for the whole tree. The build refuses to guess: in a shallow clone (the default for CI checkouts) git can only attribute every file to the single commit it has, so lastmod is omitted entirely rather than claiming the site changed wholesale on every deploy — Google discounts a sitemap that does that, which would cost more than the absent dates. Hosts that want the dates need full history; README says so. Declaring @astrojs/sitemap explicitly replaces the copy Starlight adds for itself rather than duplicating it — that's the supported way to reach these options, and Starlight's own version only configures i18n, which this single-language site doesn't use. Adds `npm run check:sitemap <origin>`: robots.txt -> index -> every advertised URL returns 200, the checks Search Console runs, against a local preview or against production. Run against production today it fails on the missing Sitemap: line, and reports the deployed site is several commits behind the repo — /concepts/permissions/ and /reference/migration/ are live 404s. |
||
|
Snow Lee (Sungwon)andGitHub
|
d362b1dbf9 | fix(cli): ask before macOS pops "Background Items Added" at init (#139) | ||
|
|
77a68542ea |
feat(security): 318 hardening fixes across hub, sync, CLI and SPA (#112)
* fix(security): close 12 authorization holes found by adversarial round 1
Four offensive agents attacked the hub's trust boundaries in isolated
worktrees; every finding is a Go test that failed on the tree before the
fix and passes after. 43 TestSec_* regression tests land with the fixes.
The two that matter most:
- A grant outlived org membership: projectPerm consulted p.Perms before
checking the org role, so removing someone from the org through the API
left their explicit project grant working. Offboarding did not offboard.
- Any device could PUT any other device's journal key — the hub never
compared the key to X-Bdrive-Device. That is the "each device writes
only its own journal" invariant, enforced nowhere.
Also: uploads accepted .bdrive/ and .git/ paths (and materialize applied
only filter.Skip, never neverSync, so a hostile peer journal could too);
the org share audit handed every member public /s/ URLs for projects they
were denied; password reset left old sessions and device tokens valid;
blobs were not verified against their content address; quota was bypassed
by chunked encoding and by a client-declared size; X-Forwarded-For
defeated both rate limiters, login brute-force included; the /s/* sandbox
CSP was missing on error paths; expired shares were revocable by anyone;
and projectPerm failed open to admin for org-less and unknown projects.
Known open, recorded in .claude/security-goal.md: the device header is
self-asserted (the fix buys an audit trail, not identity), and the
Dir==nil/Auth==nil admin escape needs a design decision first.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 17 more holes found by adversarial round 2
Round 2 attacked the five boundaries round 1 never reached (read heat,
path handling, secret leakage, the agent hook guard, the metadata store)
plus the places round 1's coverage was overstated. 43 more regression
tests; 86 TestSec_* now green.
The critical one was not on the original board. A journal op's Blob field
is a raw storage key that nothing validates: handleStorePut checks blob
keys and their content hash, but a journal is arbitrary JSONL that never
passes through that validator. One PUT of your own journal with
"blob":"../../<other-project>/blobs/<sha>" — or "../../../etc/passwd" —
then reads back through the ordinary /file, /download and /render routes.
Any member read any file on the hub host, across orgs and outside the
storage root. Guarded where content resolves, and localBackend now
refuses a key that escapes root at all.
Also: /blob served HTML and SVG inline on the hub origin with no sandbox
CSP (stored XSS via History); bdrive init deleted the hooks it had just
written whenever $HOME is a git repo, silently disabling sync hooks
machine-wide; any account could rewrite another org's device registry row
and forge History attribution; a planted device id became a heat "reader",
putting an identity in an API response that must never carry one; storage
errors relayed the hub's absolute paths (and on S3 the bucket and key);
/auth/login?next= was an open redirect via backslash and TAB; the org and
project registries handed out their live maps (self-promotion to owner,
plus a hub-killing concurrent map iteration); revoked invites came back
after a restart and refused writes applied in memory anyway; a share
minted by someone since removed from the org kept serving publicly;
a newline in a folder name made the hook guard spawn bdrive outside any
mount; single-volume upload escaped through a symlink; the seat check was
check-then-act; and the hub data dir holding auth.json ended up 0755.
Known open and recorded in .claude/security-goal.md: nothing expires (no
TTL on tokens or sessions), the Dir==nil/Auth==nil admin escape, and
client-asserted Op.User. Postgres was never exercised — row 14's
SQL-injection result covers file and sqlite only.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 17 more holes found by adversarial round 3
Round 3 attacked the journal fields nobody had audited, the fixes rounds
1 and 2 had just landed, the real `bdrive serve -c` config path, and the
read ledger on a live Postgres. 133 TestSec_* now green.
The two worst are in the syncer, and both are arbitrary file write on
every teammate's machine. materialize guarded Path with neverSync, which
splits on "/" and looks segments up in ReservedDirs — ".." is not a
member and no ignore rule mentions it, so one JSONL line pushed to your
own journal key reached ~/.ssh/authorized_keys on every device that
synced. And the reserved-directory guard was case-sensitive, so
.GIT/hooks/pre-commit sailed past it and APFS resolved it into the real
.git/hooks. Op.Mode was applied verbatim, setuid bit included.
Attacking our own round-2 fixes paid for the slot twice over: ownsDevice
turned out to be a one-request speed bump (the refused report is what
registered the forged id, so the second identical request passed) and
its first-caller-wins rule made device registration a claim-staking
primitive an outsider could use to forge History attribution and
silently kill the real owner's read heat forever. Both dissolve by
keying the registry on (account, id) instead of treating the client's
device header as a hub-wide namespace. trust_proxy, added in round 2,
took the first X-Forwarded-For hop — but XFF grows left-to-right, so
turning it on disabled the login brute-force limiter instead of fixing
it.
Also: a peer's Lamport: MaxInt64 wrapped a victim's clock and silently
reverted its own edits on its own disk; History leaked other orgs'
device names and was a hub-wide device-existence oracle; anonymous
/api/config named the storage bucket; /auth/reset enumerated accounts by
timing and was not rate limited; the signed-in hub UI was frameable and
sniffable; an asset miss returned the app shell marked immutable for a
year; Op.Size forged Content-Length; a share on an org-less project
survived its creator's offboarding; a refused upload still created
directories outside the served folder; and one NUL in a read report
wedged the entire hub's read telemetry permanently on Postgres.
Verified against a real Postgres 16 this round. Known open and recorded
in .claude/security-goal.md: NUL round-trip on Postgres text columns
(refusal vs encoding is a design call), nothing expires, the
Dir==nil/Auth==nil escape, client-asserted Op.User, and store/sign on a
backend that can actually presign — never reached in three rounds.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 32 more holes found by adversarial round 4
Round 4 opened five packages no previous round had touched — store,
journal, config, remote and cmd/bdrive — and attacked the fixes rounds
1-3 had landed. It broke five of them. 166 TestSec_* now green.
The four criticals:
- ownJournal bound the journal key to the X-Bdrive-Device header of the
same request, and nothing bound that header to an account. Round 1's
test varied the key while holding the header fixed — the one
combination that was refused. Move both together and any member wrote
and replaced any peer's journal: her ops vanish, every device replays
the attacker's deletes, History credits them to her.
- store.BlobPath joined Op.Blob straight onto the blob dir with no check
that it was a sha256, and HasBlob answering true made syncer.pull skip
hash verification — so a peer's journal op read any file on every
teammate's machine into their working folder.
- A folder's .bdrive/config.json chose where this device's hub token was
sent. The file travels with the folder, so a zip or a colleague's copy
redirected the credential to any host, http:// included.
- sync --prune read .bdriveignore before the cycle and pruned against the
version the cycle had just pulled, so it deleted for the whole team
under exactly the ! rules it refuses to run with. A teammate running
bdrive scope was enough; no attacker needed.
Also: two symlink escapes in materialize (unsafeRel judges spelling, not
disk); three ways one peer op killed sync permanently on every device
that pulled it, including a panic on a short blob string; the ignore-file
reload dropped the nested-mount boundary; round 3's Lamport ceiling was
inclusive and so still reachable; the (account, id) device rekey held
neither on the read path nor on disk, so a restart handed the device to
the squatter; presigned uploads bypassed the content-address guard
entirely — that whole branch had never executed under a test, since every
fixture used file:// which cannot sign; remote.Prefixed, the single
containment primitive for multi-tenancy, did not contain; and a mount id
from the untrusted folder config escaped $BDRIVE_HOME.
Plus: bdrive export wrote hub-named keys as tar members unvalidated; the
device token followed cross-origin redirects; a symlink in the file://
storage root read and wrote anywhere on the host; GCS presigned PUTs
bound no size; presigned device uploads were never billed; trust_proxy
was wrong for the third round running; one bad line voided an entire
journal; Op.Path was not byte-exact through JSON; Less was not a total
order, so Replay's determinism rested on a caller's accident; and client
journals were 0644.
Three new scoreboard rows: client local state, the project archive, and
the device as client of a hostile hub. Still zero tests after four
rounds: internal/daemon and internal/autostart.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 29 more holes found by adversarial round 5
Round 5 opened internal/daemon and internal/autostart (zero tests after
four rounds), drove the browser presign flow, and attacked round 4's
fixes. It broke seven of them, including the one round 4's commit message
called "the critical". 247 TestSec_* now green.
ownJournal failed four separate ways. The write doors called
observeDevice BEFORE ownJournal, so for an unclaimed id the request
manufactured the fact that authorized it — and device ids are not secret,
History publishes them. The claim was then permanent and unrecoverable:
no Delete on DeviceRepo, no release, no admin route, no CLI re-mint, and
a 403 naming no remedy, so any member could deny any colleague's laptop
the ability to sync forever. Rows were global but visibility per project,
so offboarding a teammate released her journal to whoever was left.
And because an ownerless row ranked as the earliest claim, every hub
upgraded from before rows had owners had the binding switched off for
precisely its established devices. Ownership is now hub-wide, an
ownerless row authorizes nobody, observation happens after the decision,
and project admin is the documented recovery path.
The other critical is a divergence primitive: pull resumed at an op
COUNT, and round 4 had just taught Parse to skip bad lines silently. One
undecodable line inserted among lines a device already counted shifts
every appended op down by one, so two devices replaying one journal hold
different states permanently — and the peer picks the split. Now resumed
at a byte offset.
Also: a mid-run edit to .bdrive/config.json moved a whole project to a
remote of the writer's choosing, with no restart and no credential, and
the daemon then pulled from it; verify-on-read was defeated by uploading
honest bytes first, since a presigned URL is replayable for its TTL;
appendOp's lamport wrapped int64 and silently broke last-writer-wins for
every later upload in the project; peer journal strings reached bdrive
log's terminal unescaped, so the audited party could rewrite the audit
(OSC 52 to the clipboard, \r to repaint a delete as a put); path_raw let
one journal line name two different files to two reader versions; Stop
signalled whatever pid a 0644 file named; locked() failed open, so status
lied and stop stopped nothing; and a macOS path containing "&" made the
autostart plist unparseable while Install reported success.
Two four-round deferrals are now answered rather than carried.
Dir==nil/Auth==nil is not reachable — nine real configurations, both
arms, real project ids — and is a guarded invariant. The Postgres NUL
question was swept across seven stored-record surfaces on a live
Postgres: no silent-loss path, and cleanUploadPath now refuses control
characters so it is unreachable through the API.
Quota became a reservation with reconciliation: reserved at the grant so
concurrent grants cannot oversubscribe, charged on arrival, released free
on expiry, and charged once.
Recorded, not hidden: permHub builds a hub with Devices == nil, so round
4's ownership binding was inert in that fixture and earlier "clean"
results measured through it proved less than they looked.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 30 more holes found by adversarial round 6
Round 6 attacked round 5's fixes, the completeness floor (templates, CLI
commands, the untested exported API), and — new this round — the test
suite itself. 290 TestSec_* now green.
One hacker reverted 33 of the accumulated fixes one at a time to see
which the suite actually caught. 28 held. Five did not, and those five
are the most valuable result of the round: permHub built its hub with
Devices == nil, so a dozen journal-pushing tests had been proving
permission and never ownership; the ownerless-legacy-row test passed
because its helper never set Op.Device, so the r5 hole it guards would
have gone undetected; nothing asserted that an outstanding presigned
grant counts against the quota cap, which is half of reserve.go's
contract; row 6 claimed cleanUploadPath refuses control characters and
named no test, and that guard is what keeps the Postgres NUL divergence
unreachable; and unsafeRel — round 3's headline client fix — could be
deleted with the suite green, surviving on round 4's UnderRoot. It also
accepted ".", contained only because hashFile happens to fail on a
directory first.
The criticals: round 5's byte-offset pull resume was the same divergence
primitive it replaced, twice. A peer that publishes in two stages and
cuts the first mid-line makes one chosen device permanently skip the op
that straddles the cut while every other device applies it. And round 5
deleted the shrink guard, so a peer withdraws an op every device already
applied — the file vanishes from teammates' folders with no delete op,
nothing in the journal, nothing in History. Separately, Deny removed an
account but every authorization decision downstream keys on email, so
grants and org roles stayed attached to the address: re-registering it
walked back in as project admin, and its public share links kept
serving.
Also: reset and verification mail took its link host from the request, so
an unauthenticated stranger could have the hub mail a victim a genuine
reset link pointing at the attacker's server; a refused password reset
reported "Password updated"; share revocation, approval, policy and
account removal all took effect in memory after the store refused them,
each failing in the widening direction; one journal push with a year-2300
timestamp overflowed the History cursor and hid the whole audit feed past
page one; account ids were 32 bits with no uniqueness check, and the
birthday bound is ~9,300 accounts for a 1% chance of silently
transferring one account's credentials onto another; the new reservation
ledger had a data race on the billing path, a check-then-act that let 5
of 16 concurrent grants through a cap fitting one, and released arrived
bytes unbilled on expiry; safeField stripped C0 but not the 8-bit C1
controls that are CSI/OSC/DCS/NEL in any xterm-lineage terminal, nor
bidi overrides; and internal/templates — first contact — bypassed
cleanUploadPath entirely and wrote through symlinks using the shipped
template with no hostile input at all.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 29 more holes found by adversarial round 7
Round 7 drove `bdrive init` end to end for the first time — named the
largest gap by two consecutive CISOs — and tripled the sabotage sweep to
53 reversions. 326 TestSec_* functions green.
init held two criticals the moment it was actually run. `init --server`
took the no-auth branch when a server answered {"auth":{"enabled":false}}
and rewrote settings.Server without touching settings.Token — and
settings.Server is the entirety of round 4's token binding, so a 30-line
HTTP server collects the real hub's bearer token. And `init $BDRIVE_HOME`
was accepted, because the .bdrive reserved-directory rule only applies to
segments below the mount root: from there settings.json is an ordinary
top-level file, so the first cycle pushed this device's token to the hub
as project content, for every member and every teammate's disk.
Two more criticals were old fixes on doors they never covered. Round 4
bound the device token to settings.Server's origin in remote.deviceToken
— the sync backend's door — while share.go reads its destination from the
folder's .bdrive/config.json and hands the token straight to it, and the
CLI's own http.Client had no CheckRedirect at all. And round 6's offboard
only log.Printf'd RemoveMember's "cannot remove the last owner", so
anyone signing up on a removed sole-owner's address inherited org
ownership and admin on every project in it.
The sabotage sweep found 8 more guards deletable with the whole suite
green, including MayActAs (every existing test planted an id that
validDeviceID rejects first, so the ownership loop was never consulted),
both framing headers (round 3's test held the disjunction, not the code),
and sqlAccountRepo's id guard — where the untested backend is the one
managed and Postgres deployments run. For the first time the sweep also
covered the three choke points themselves: reverting requirePerm turns 30
tests red, projectPerm 21, authGate 9.
Also: two more journal-undo primitives past round 6's count guard, now
keyed on identity; reset-mail poisoning survived because the pin was
first-request-wins and round 6's own reproducer sent the honest request
first; the /store/* journal door accepted paths /upload/commit refuses,
so the three spellings of one path rule became one exported predicate;
bdrive forget injected .bdriveignore rules outside any managed block;
bdrive resume built a volume path from an unvalidated registry key; round
5's $HOME-is-a-git-repo fix broke again on a string compare of two
spellings of one path, silently disabling every agent hook on the machine
while init reported success; and the CheckWrite call round 6 moved under
the hub-wide ledger mutex stalled every project's sync cycle.
Known open: a peer can still un-publish an applied op by corrupting its
line and appending as many as it removed — the clean close is hub-side
append-only enforcement on /store/*, a behaviour change no failing test
demands yet. The init tests all ran with auth.enabled false, so the login
flow inside init — where the first critical lives — is still untested.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 26 more holes found by adversarial round 8
Round 8 drove the authenticated login flow, the last five CLI commands,
and sabotaged row 15 exhaustively. 621 TestSec assertions green.
Three of round 7's fixes were half-fixes. Its $BDRIVE_HOME guard closed
the child direction only, so `bdrive init` on any ANCESTOR of the home
still pushed this device's bearer token to the hub as project content —
and a relative $BDRIVE_HOME disabled the guard entirely, reopening the
original critical. Its mail fix still seeded the pin from r.Host, so on a
fresh process an anonymous stranger could have the hub mail a VICTIM a
valid reset link on the attacker's server; the mirror image turned
password recovery off hub-wide with one request. And the hole round 7
declined to close got a reproducer that also invalidated its proposed
remedy: a peer publishes its last op unterminated, the next append fuses
onto that line and stops it decoding, so an op every teammate has on disk
vanishes with no delete op — through a byte-level PURE APPEND, which
hub-side append-only cannot stop.
Two more criticals were first contact. The loopback login callback had no
proof of possession: its only binding is a `state` that is printed to
stdout and passed to xdg-open as argv[1], so any local process that can
run `ps` signs the device in as its own account and the user's folders
sync into the attacker's project. And the three /store/* READ doors call
observeDevice as their first statement — round 5 moved it after the
decision on the write door and never touched the read doors — so one GET
with a victim's device id first-claims it hub-wide and locks that device
out of its own journal, from read permission on any single project.
The sabotage sweep is the round's most important result. 48 guards in row
15 reverted one at a time: only 20 were caught. A 57% false-negative
rate, nearly 4x rounds 6 and 7. The materialize DELETE loop's three
guards — one of which ends in os.Remove — were held up by nothing,
masked because scan's delete pass applies the same rule to the same cache
first, so no whole-Cycle fixture can tell which guard refused.
absorbLamport's ceiling and tickLamport's stop mask each other, so the
existing test passes with either removed. Eight of the 26 misses now have
tests; 18 remain open, and row 17 was never reached at all.
Also: a hostile export archive chose which of your existing projects it
landed in (create-or-join-by-name, emptiness checked after the join); a
folder that merely arrived on disk stole an enrolled mount's registry
row, so at next login the real project's daemon ran on the arriving
folder; one device approval minted N tokens and bound them to a device
the human never approved; logout left the credential live with no
revocation route; and two unbounded reads on the device side had the
declared size in scope at the call site.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 5 more holes found by adversarial round 9
Round 9 spent most of its budget on sabotage rather than new attacks, and
the numbers are the point. 705 TestSec assertions green.
Row 17 swept in full: 44 reversions, 9 missed (20.5%). Rows 13 and 18-22:
57 reversions, 12 missed (21% overall — row 13 at 40%, row 20 at 33%,
row 18 at 0%, the only perfect row swept in any round). And row 15's 18
remaining untested guards got tests, each verified red under its own
reversion — including both of round 8's flagged leads, which turned out
to be correct guards that were merely untested. That agent reported dry.
So roughly 29 previously-deletable guards are now genuinely pinned.
All five live holes are regressions in round 8's own fixes. Its
re-assertion fix re-published a withdrawn op at its original low lamport,
which made it a losing local unpushed op the instant it was written — so
conflictCopies did what it exists to do and the victim created, signed
and pushed a file holding content the peer chose, at a path that never
existed. The same admission rule had no guard for withdrawn deletes and
consulted neither the ignore filter nor neverSync, so a device
republished paths it deliberately refuses to materialize. Re-assertion
now requires that this folder's own cache stands behind the op.
Its sizeBound fix returned on a sha mismatch instead of skipping, so one
understated Op.Size in one line of a peer's journal permanently withheld
every blob queued behind it — before round 8 the read was unbounded, the
sha matched, and the files arrived. The bound was not the defect and is
unchanged; the error is now remembered and returned after the batch.
Also: the org heir was chosen by the smallest email address rather than
the longest-standing member, so the newest member inherited ownership and
project-admin on every project when a hub admin removed a departed
employee; $BDRIVE_HOME was created 0755 by LoadDevice, which runs before
almost everything, so listing alone named every project, every device in
the fleet and every content hash without opening one of the 0600 files;
and round 8's ResolveMount condition stranded a genuinely moved project
behind a leftover config, with init itself blocked on the same check.
Move-vs-copy is now decided by dev+ino, which a rename preserves and a
copy cannot reproduce.
Judgement call recorded: the nested-mount carry reverts green but stays.
Deleting a defence-in-depth guard because the tests did not notice is
exactly the reasoning the sabotage table exists to distrust. It is marked
as not counting toward coverage.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 30 more holes found by adversarial round 10
Round 10 drove three surfaces end to end instead of sweeping, and the
contrast is the round's real result. Round 9 spent itself on sabotage and
found 5 holes; round 10 pointed a device at a hostile hub, executed the
Linux autostart code for the first time in ten rounds, and requested the
three /auth pages nobody had ever fetched — 30 holes. 777 TestSec
assertions green.
Row 19 is the clearest data point the loop has produced. A round-9 sweep
scored it 12.5% missed and annotated it "no reachable impact", so it
wrote zero tests. Driven end to end it held 11 holes, one critical: pull
skips a listed journal only on an exact string compare, so a hub listing
journal/DEVA.jsonl resolves to the same file as deva.jsonl on APFS and
NTFS and overwrites the device's OWN journal — the invariant the whole
concurrency design rests on. A sweep can only find a hole where a guard
exists; it cannot find a class nobody wrote a guard for, and every
critical since round 7 has been that kind.
Also from the hostile hub: one unusable listed key hid every peer
permanently; one listing minted 200k local journal files; the hub sized
the device's own allocation at two layers, and one boolean in a sign
response made a device publish an op for content it never sent, with the
cursor advanced past it so it never retried. putDirect shipped file
bytes to any host the hub named — round 4 dismissed this because "the hub
already holds the data", but at the moment it names the destination it
does not, which is what the upload is for.
Row 5's device binding is closed after four rounds of deferral. A
read-only member's device could never register, so any member with write
anywhere took its id permanently — and the arm that let them through read
a field the attacker writes. The id is now minted hub-side at login,
bound to the authenticated account, at all three mint points. That was
first framed as requiring the supersession of round 7's test; it did not.
Round 7 asserts a read door creates nothing, and that property is
unchanged and strictly stronger, because the read door now has nothing
left to claim with.
Four of round 9's five fixes had live residuals: re-assertion laundering
returned on any device that cannot push (conflictCopies measures unpushed
against a cursor that only advances on success, and read-only is the
documented steady state); the sizeBound fix still let one peer integer
suppress the victim's own push; earliestMember was inert on every
upgraded hub; and the dev+ino discriminator was inert on every row that
existed. ResolveMount turned out to be a write with a read-shaped name,
so bdrive restore and forget enrolled the device in projects it was never
init'ed into.
Recorded as a measurement gap, not a finding: row 14 was scored clean on
every backend for seven rounds, but this is the first round ever run with
a Postgres DSN, and metaBackends silently omits the arm without one. A
skipped arm and a missing guard are indistinguishable in a green suite.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 24 more holes found by adversarial round 11
Two criticals, both first-contact findings on surfaces ten rounds had not
reached.
An `.xml` file got script on the hub's own origin with the reader's session:
`sandboxInline` walled off a LIST of content types where the thing it protects
is a PROPERTY ("the browser parses this as a document"), and the whole XML
family sat outside the list while having the property — an XML document carries
its own `<?xml-stylesheet type="text/xsl"?>` and the XSLT output is HTML in the
origin that served it. `inlineMarkup` is now the property, `inlineType` serves
the XML family inline as text/plain so nothing parses it as a document, and
`nosniff` goes on every stored-bytes door.
A plain member replaced another account's journal by spelling her device id in
a different case: every hub ownership decision was a byte compare while APFS
and NTFS fold, so one login and one PUT broke the one-writer invariant the
whole concurrency design rests on. `canonDeviceID` folds at the trust boundary,
the registry folds on load and at every entry point, and `ownJournal` requires
the canonical journal key.
Also: org ownership was drawn by Go map iteration (`sort.Slice` on an all-zero
`Created` column) and now needs real evidence of age or produces no heir; a
revoked grant was restored by any unrelated write from a second hub process
(grant writes are row-scoped now, on all three backends); `/history` named
whoever the pushing device typed; the admin recovery arm locked the real owner
out of `bdrive login` forever, across the org wall; `bdrive scope` was the
unescaped door `forget` used to be and could wipe the team's synced rules;
`journal.SafePath` let every bidi control and every C1 through.
Two tests were touched, both disclosed in .claude/security-goal.md:
TestSec_DB_NULBytesDoNotTruncateRecords is retired (its assertion is one
Postgres cannot implement), and TestSec_Scope_AddCannotCreateADirectoryOutside-
TheProject is rewritten against the guard it was meant to test — it called
os.MkdirAll itself, so no production code sat between its setup and its
assertion. The rewrite was verified to go red with the guard removed.
918 TestSec assertions green, whole suite green including Postgres and -race,
108/108 Playwright, Linux container clean.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): the run-mode note only printed when the suite was already red
Round 10 moved the "postgres UNTESTED in this run" note from t.Log to
os.Stderr because t.Log is invisible without -v. That was not enough: `go
test` buffers a package's output and discards it on success without -v,
stderr included. So the note that exists to make a silent coverage gap loud
was itself audible only during a failure — the same shape as the hole it
guards against, and the reason it went unnoticed is that every round that
read it had a red suite in front of it.
secrunNotify now also writes to /dev/tty, which survives that buffering.
Verified under a pty on a fully passing run with no -v.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 24 more holes found by adversarial round 12
The critical one is the fifth instance of "something survives offboarding":
an org invite outlived the membership, the ownership AND the account that
minted it. OrgDB.Redeem/ValidInvite now resolve the minter's ownership at
read time — the rule shareCreatorStillBelongs already applied to a share
link — and retire the invite when it fails, so EvictMember's heir promotion
cannot revive it.
Round 11's row-scoped write landed on ProjectRepo only. OrgRepo, ShareRepo
and DeviceRepo had the same whole-record shape, so a second hub process's
unrelated write resurrected a revoked org membership, a revoked /s/ link,
or erased a device binding. And the authorization READ path never got it at
all: ProjectDB answered from a copy taken at boot, so a revocation took
effect on one process and no other.
Also: a push could credit another account through Op.Author; journalOps
checked the Note and not Author/UserName; a display name skipped trimText;
a password reset left outstanding reset and verification mail grants alive;
nosniff missed two stored-bytes doors; SafeText admitted the zero-width
formats; a project name could break out of the ConnectGuide paste prompt;
inviteTokenFromNext matched "/join/" anywhere in `next`; file content chose
what read-log reported as a read, and the hub recorded reads for paths that
do not exist.
Frontend: decodePath threw URIError on a link in a teammate's document and
unmounted the whole SPA persistently (fixed at the decode, plus a real
ErrorBoundary), and the router kept the Object.prototype lookup bug round 11
fixed in ProjectIcon.
Decision, not a patch: agent HOOK config (.claude/settings.json and
friends) is now reserved in both directions. Skills, commands and CLAUDE.md
deliberately are not — sharing what an agent reads is the product. The trust
boundary that follows from that is now written down in INSTALL_FOR_AGENTS.md,
the docs' Start-here path, and README.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 17 more holes found by adversarial round 13
Round 12 gave ProjectDB a read-path refresh() because a revocation only took
effect on the process that served it. Four sibling registries with the
identical defect went unexamined for a round.
The refresh family (one fix, three service structs + two file repos):
- OrgDB.refresh — the wall IN FRONT of project permissions. A removed org
member kept reading every project in the org; a revoked invite still
redeemed and, on the default invite-only posture, bootstrapped the account.
At the top of the MUTATORS too: the last-owner guard is a cross-process
TOCTOU the write-side re-read cannot close.
- BuiltinAuth.refresh — the CREDENTIAL, not a grant on top of one. A revoked
device token still authenticated; a deleted account signed in again with its
old password.
- ShareDB.refresh — a revoked /s/<token> was still served to anonymous
strangers. fileShareRepo.reload's own r12 comment named this row.
- fileAccountRepo.reload, fileReadRepo.reload — the last two file repos with
no write-side re-read.
It lives in the service structs, not db_file.go: the staleness reproduces on
sqlite and Postgres too, and a file-only fix would have left the
two-replicas-one-database deployment fully broken.
Also:
- offboard now releases the device binding (DeviceRepo.Delete +
DeviceRegistry.Release). A deleted account kept a hub-wide claim on its
device id, which silently and permanently locked out the next hire.
- SafeText and trimText refuse unicode.Cf and the tag block AS A CLASS. The
tag block encodes all printable ASCII with no glyph, so a project name
rendering as "wiki" smuggled a shell command into the agent paste prompt.
- .mcp.json is reserved; the agent-config list is now derived from what each
platform LOADS, not from what BearDrive writes.
- pageDevice and X-Bdrive-Device-Name go through trimText: an unauthenticated
stranger chose the text AND the length of the hub's only consent surface.
- SetPolicy runs the startup validator, so POST /api/admin/policy cannot reach
a posture the binary refuses to boot in.
- Insights uses Object.create(null): a folder named __proto__ erased an agent
device from the Dashboard.
go build / go vet / go test ./... clean with and without BDRIVE_TEST_POSTGRES;
-race clean; Playwright 127/127. Linux container run not completed — see
.claude/security-goal.md.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close the agent-onboarding holes found by round 13's fourth hacker
Four findings, all against the already-hardened tree. Its fifth (.mcp.json)
was already closed and its test passes — an independent confirmation of that
fix from a second angle.
1. A peer's `!.env` in the shared .bdriveignore uploaded another member's local
.env on their next cycle. Round 4 made .bdriveignore team-wide on purpose
and made `sync --prune` refuse on `!` rules for exactly that reason — but
that reasoning covered DELETION, and nobody asked what a pulled negation
does to SCAN. The runbook's own `bdrive init . --only docs,notes` is what
creates the exposure: the whole repo goes under the mount with only this
synced, teammate-writable file holding the rest back.
Fixed asymmetrically at the upload door (Filter.SkipUp, consulted by
walkFolder): pulled rules that NARROW apply immediately in both directions;
pulled rules that WIDEN apply to materialize but not to scan, until this
device authors the rules itself (init --only, bdrive scope, an editor). A
joining device has authored nothing, so team-wide scope still works on day
one — which a blanket "ignore pulled negations" would have broken.
`bdrive scope --explain` reads the same floor so it cannot drift.
2. A FAILED `init --server <url>` signed the device out of its real hub and
left it defaulting to the new one — after a run that ended in "Error:".
ensureLogin now returns a rollback; initCmd commits the session only once
the hub has answered with a project this device can open.
3. `init --server http://…` minted and stored a device token with no plaintext
warning while `bdrive login` on the same URL warned — and step 2 of the
runbook is titled "Do not run a login command". The warning moved from
loginCmd's RunE into the shared runLogin: one sign-in door, one warning.
4. The hub chose the device-login link and the CLI printed it verbatim under
its own "open this link in any browser". sameOriginLink falls back to the
hub's own /auth/device when scheme+host differ.
Also: safeField gets the same unicode.Cf + tag-block class rule SafeText and
trimText got — third door, same class. Scoping guide documents the widening
rule. Two findings that are not tests (the runbook URL pinned to a mutable
branch; nothing authenticates the hub during device sign-in) are recorded in
known-open.
go build / go vet / go test ./... clean with and without BDRIVE_TEST_POSTGRES;
-race clean on webapp, syncer, store, cmd/bdrive; Playwright 127/127. Linux
container run still not completed — the Docker daemon on this machine will not
create containers at all; see .claude/security-goal.md.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(security): close 22 more holes found by adversarial round 14
The last round of the loop. Four hackers landed; 21 of their tests go green
here, the 22nd names a hole that IS fixed but cannot pass as written (see
below). `.claude/security-goal.md` gains a handover section for a human in
place of the "next round's targets" framing.
Sync / scope (internal/syncer)
- A peer DELETING the shared `.bdriveignore` walked straight past round 13's
upload floor: the `IgnorePulled` bookkeeping sat behind `if want, ok :=
target[IgnoreFile]; ok`, so a delete never updated it while materialize
unlinked the local copy anyway — and the next cycle read the absent file as
locally authored, dropped the floor AND the live rules, and pushed the whole
repo. Recorded as pulled now.
- Both SyncState fields are `omitempty`, so on its first post-upgrade cycle
every existing device adopted whatever was on disk as its own — including a
peer's `!.env` that landed one cycle earlier (scan runs before pull, so it
always does). An upgraded device now seeds its floor with `vouchedFloor`:
keep a `!` line when the path it re-includes is already in this mount's
materialization cache, drop it otherwise. That is what keeps a `bdrive
scope` block (`/*` plus nothing but negations) from silently ending uploads
— pinned by TestUpgradedScopedDeviceKeepsUploading.
Second-process staleness (internal/webapp)
- ProjectDB had refresh() on Get and List only. `put`→`PutMeta` is an
unconditional upsert, so a second process's ordinary rename put a DELETED
project back carrying the org the public-link rule reads; the last-ADMIN
guards counted admins out of the boot-time map; GetOrCreate answered
create-or-join differently per replica. refresh() now runs at the top of all
ten mutators.
- DeviceRegistry had no refresh() at all — round 13 cleared it on the
bind-away direction alone. Offboarding released a device claim on one
process and no other: the next hire is locked out, and a re-created address
inherits the departed account's journal write gate elsewhere. Pinned on
file, sqlite and postgres.
Audit trail
- `/store/object` was a plain object PUT with no relation to what is stored,
so any member could rewind their own journal — or, after inheriting a
reassigned device id, a departed member's — out of History. Journal pushes
now must keep every op Seq the hub already holds.
- seedTemplate journaled the hub's own template files under the account that
ran `bdrive init --template`, byte-identical in shape to a hand upload. They
now carry no account and a "seeded from the <name> template" note.
Text and rendering
- journal.SafeText refused every category-Cf rune and missed U+2028/U+2029
(Zl/Zp), which the webapp's own trimText has deleted by number since round
12. A folder row for `line<U+2028>sep.md` paints to exactly the same glyph
run as `line sep.md`.
- A strong-RTL LETTER needs no format character to reorder a rendered row.
Measured in Chromium, `unicode-bidi: isolate`, `plaintext` and `<bdi>` all
leave it intact; `isolate-override` fixes it, and peer-written-name
selectors now carry it (SPA + the auth pages' device-approval rows).
- `)` closes the paste prompt's clause exactly as `"` did; project names now
drop both parens (org/device/account names are unaffected). PATCH
/api/projects/{id} called trimText where create called trimName, so rename
stored `/` and `\` — one rule for both doors now.
CLI
- `p.Template` was the one hub-chosen field in `bdrive init`'s output that
never reached safeField.
- `--template` reported the hub's own string as proof and never looked at what
arrived; it now always falls through to the idempotent seedLocally.
Docs
- INSTALL_FOR_AGENTS.md no longer raises the hub-seeded AGENTS.md to the
user's authority, and its trust boundary names the hub as an author of
folder content. Documentation defects, not demonstrated exploits: three live
headless runs did not flip behaviour.
Known red, deliberately not worked around:
- TestSec_ProjectName_RenameBypassesTheCreateNameRule — the hole is fixed and
separately verified; the test's own control creates a project with the
normalized name in the same org before renaming into it, so correct
behaviour collides with the unique-name-per-org rule.
- Two e2e/sec14fe specs upload a U+2028 path to demonstrate a rendering
collision that this commit's ingest fix now refuses at the door.
go build / go vet clean. go test ./... green with and without
BDRIVE_TEST_POSTGRES and under -race, except the test named above. Playwright
serial: 131 passed, 2 failed (the two named above).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* test(security): the rename test collided with its own control
TestSec_ProjectName_RenameBypassesTheCreateNameRule created a project
from the same payload it then renamed a second project to. Both
normalize to "notes....etc", names are unique per org, so the rename
400'd on the collision and the test failed at its control check without
ever reaching the assertion it exists to make.
It failed that way against the FIXED code, which is the worst way for a
test to be wrong: it reads as an open hole and is really a broken
instrument. The control now uses a payload that normalizes to a
different name.
Verified load-bearing rather than merely green: reverting Update's
projectLabel call turns it red with "rename stored a path separator in
a project name: notes/../../etc".
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* docs(architecture): draw the security hardening's new types and seams
The branch added shared rules and gates that the diagrams did not have
boxes for, so the pictures no longer described the code.
cli-sync: the three drifting path/text checks collapse into one
journal.SafePath/SafeText box, plus store.UnderRoot and
config.ReservedPath as their own single-rule boxes; Filter gains SkipUp
and AcceptRules and the scan/materialize rules stop being symmetric;
SyncState is drawn for the IgnoreAccepted/IgnorePulled floor; the
registry records Dev/Ino and splits ResolveMount (read, self-heal) from
EnrollMount (the only writer); the daemon's signalled pid moved inside
the flock.
webapp-server: DeviceRegistry is keyed (account, id) with
Bind/Release/OwnerOf/MayActAs; new boxes for the /store journal door,
the quota reservation ledger, sandboxInline and offboard; refresh() on
every service; the row-scoped repo interfaces, the storable validation
gate and the schema-version guard in the MetaStore block; HasBlob became
BlobSize.
webapp-frontend: ErrorBoundary, the app's floor. Also fixes escaped
quotes in a note that have been rendering this whole diagram as an error
box on GitHub since it was written.
overview: unchanged — no package appeared or disappeared and no
cross-piece flow moved.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* test(security): re-aim two specs their own fix made unbuildable
Round 14 refused U+2028/U+2029 at ingest. Two e2e specs proved a
*rendering* collision by uploading exactly those code points, so the fix
made their fixture impossible and left them red — a broken instrument
reading as an open hole, the same shape as the rename test.
The listing spec now asserts the ingest guard instead, and keeps the
measurement that says why the guard matters: the two names painted to
70.9844 x 16, one line box, byte-different and pixel-identical, measured
in Chromium with Range.getClientRects() over live text nodes. Relax
SafeText and it goes red before the collision returns.
The shares-audit spec is skipped with its numbers preserved. Its
reachable sibling — a strong-RTL letter, which cannot be refused without
refusing Hebrew filenames — is covered by the neighbouring spec and
fixed in style.css.
Playwright: 133 passed, 1 skipped.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* test(security): a comment in a spec leaked a CSS rule into the bundle
Tailwind's content scanner reads the e2e specs. A bare `isolate` token in
prose explaining the bidi fix emitted `.isolate{isolation:isolate}` into
the shipped bundle — a rule nothing uses, and a stale-assets failure for
check-dist.
Reworded to avoid the token, with a note saying why, since the next
person to explain a utility class in a comment will hit the same thing.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* docs(security): correct three handover entries the landing pass closed
The handover listed one Go test and two browser specs as permanently red.
All three were re-aimed after the loop stopped: the Go test was failing
against the FIXED code because its own control collided with it, and the
two specs were fixture-blocked by round 14's own ingest fix.
Suite is 1052 TestSec assertions green, 0 red; Playwright 133 passed,
1 skipped. The stopping condition itself is still not met and the notice
at the top of the file says so.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(sync): a Persian filename permanently wedged a device's sync
The scan door applied config.ReservedPath; the hub's ingest door applied
journal.SafePath AND config.ReservedPath. A file whose name the scan
accepted and the hub refused was blobbed and journaled locally, and
because push PUTs the whole journal object, every later push from that
device 400'd on the same op forever. Renaming did not help — the delete
op names the same path. The only surface was a line in daemon.log, and
recovery meant deleting the volume store.
walk.go's own comment already stated the rule it was breaking: "the
outbound half has to match the inbound one." It just did not name every
predicate the inbound half applies.
Two changes:
SafePath now permits ZWNJ and ZWJ, which the Cf class rule refused. Both
are orthographically required — U+200C is what makes "میروم" the right
word in Persian and is mandatory in several Indic scripts, and U+200D
builds most multi-person emoji. Refusing them did not harden a hub; it
told those users their filenames were illegal. The confusability they
buy is also already reachable without them: a Cyrillic homoglyph
produces the identical "two rows, one reader" tree and is allowed. So
the clause was paying a hard i18n cost for a partial mitigation of a
class that stays open. A note has no orthography, so SafeText still
refuses all four zero-widths; the two rules now share one implementation
with a flag rather than diverging.
walkFolder applies SafePath, so an unsyncable name is simply not carried
— it shows up in `bdrive scope --explain` like any other exclusion and
never enters a journal.
The regression test asserts a PROPERTY over both predicates rather than a
character list, so adding a rule to one door alone reopens it. Its first
draft was vacuous — it skipped every unsafe name instead of creating it,
so nothing hostile reached the scan and deleting the guard changed
nothing. Caught by sabotaging it; it now creates the names unix permits
and goes red when the guard is removed.
Found by a CTO review of PR #112, not by the security rounds, which
checked this asymmetry on the hub and never on the client.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(hub): a proxied hub throttled all its users as one on day one
Upgrading a hub behind nginx / Caddy / Fly / Cloud Run without editing its
config made every user share ONE 10/min login bucket and capped public share
links hub-wide at 120/min: clientIP fell back to r.RemoteAddr, which is the
proxy. Correct passwords started answering "too many attempts", with no log
line saying why.
X-Forwarded-For is now trusted by PEER rather than by configuration: a proxy
that fronts a hub reaches it over loopback or a private address (sidecar,
container network, Fly/Cloud Run internal hop), so that header is the
operator's own infrastructure. A hub on a public IP still ignores it, and now
logs once instead of failing silently. trust_proxy remains the override for
the one shape the peer check cannot see — a proxy on a public address.
Which hop is taken is unchanged (last element of the last field line), and the
round 13/14 tests that pin it stay green.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* perf(hub): authorization stopped re-reading the whole registry per request
Every authorized request re-read the entire project, org and account registry
under a hub-wide mutex — 14.1 ms per request at 5k projects on the file
backend, nine unfiltered SELECTs on Postgres — so the hub served roughly 60
req/s regardless of cores while ~200 devices polled /store/list every 10s.
None of this existed before the security rounds; registries loaded once at
boot.
The semantics do not change. Registries still re-read the store before every
authorization decision — that is the correctness floor rounds 12-14 built, and
a TTL would put back exactly the staleness window they closed. What changes is
the cost of asking whether there is anything to re-read:
- new optional Versioned repo capability: one os.Stat (file) or one
primary-key lookup on a per-registry meta_version counter bumped inside
every write transaction (SQL). A repo that cannot answer is treated as
changed, so the fallback is the unconditional re-read that was always there.
- proj() resolved the project and then projectPerm resolved it again;
projectPermOf takes the Project the choke point already has. handleProjectList
and the org share audit did one resolution PER PROJECT in a loop; both now
pass the row they are already holding.
Measured on M1, benchtime=200x, one project resolve + permission check:
file 100 projects 331 us -> 6.0 us
file 1000 projects 2.85 ms -> 3.9 us
file 5000 projects 14.14 ms -> 3.9 us
sqlite 100 projects 295 us -> 22.9 us
sqlite 1000 projects 2.57 ms -> 22.5 us
sqlite 5000 projects 10.86 ms -> 21.5 us
and it is now flat in project count rather than linear.
The file backend does NOT become multi-process-safe from this: every write is
still read-modify-write-rename, and the mtime+size token would miss two
processes writing the same byte count within one filesystem timestamp tick.
refresh narrows the stale-read race; it does not close it. SQL is the fix.
TestVersionGateSeesAnotherProcessWrite pins the property that matters — a
second process's create and grant change are both visible through the gate, on
file, sqlite and Postgres. BenchmarkRegistryRead / BenchmarkAuthorizedRequest
go linear again if this regresses.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* perf(hub): every blob read on S3/GCS paid double egress and a full hash first
RemoteSource.verify re-read and re-hashed the whole object before returning a
second read for the actual stream, on any backend that can presign — and
PutSigner survives the Prefixed wrapper, so this was live on every S3 and GCS
hub. Every viewer open, render, download and /s/* share hit paid 2x
object-store egress and a serialized full-object hash before the user's first
byte. On file:// (the OSS default) verify is a no-op, which is why the suite
never felt it.
The check is not weakened. It is cached again, keyed on the one thing that
makes "blobs are immutable" true rather than assumed: BOTH presign doors
refuse to sign a key that already exists, so every presigned URL a blob ever
gets was minted before its first PUT and dies at mint+TTL. Once the stored
object is older than the presign TTL, no live URL for it can exist and none
will ever be minted again — the hub is the only writer left, and the hub hashes
what it relays. Only then is the verification cached. The object's age is read
after the hash, so a replay mid-check reads as seconds old and is not sealed.
Measured, 4 MiB blob, 200 reads (the "unsealed" arm IS the old code path):
before 2.41 ms/op 2.000 storage reads per blob read
after 0.39 ms/op 1.000 storage reads per blob read
On S3/GCS the second read is real egress and real latency, so the win is
larger there than this local stand-in shows.
NOT done: signing the content hash into the presigned URL. GCS cannot bind a
SHA-256 at all — x-goog-hash takes only crc32c and md5, and the md5 would be
declared by the same client that declares the sha, so a chosen-prefix collision
defeats it. On S3 the SDK hoists ChecksumSHA256 into the query string rather
than into SignedHeader; it is inside the signature, but whether S3 enforces a
hoisted checksum (and whether an unsigned request header would override it)
cannot be verified without a real bucket. With the seal in place the checksum
would add no security that verify is not already providing during the only
window it applies to, so it stays out rather than going in untested.
Backends that can sign but cannot bind a content address: GCS certainly, S3
pending a live check.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* docs: record the three PR #112 cost fixes and what they do not fix
architecture/webapp-server.md gains Versioned/versionGate, RemoteSource's
PresignTTL + seal, remote.Object.Modified, and the new clientIP peer rule.
Both mermaid blocks parse-checked with mmdc.
.claude/security-goal.md's "known-open, deliberately deferred" list is updated
with what each fix accepts: the private-peer widening on X-Forwarded-For, the
fact that the file backend does NOT become multi-process-safe from the change
token (read-modify-write-rename is unchanged; mtime+size narrows the race and
does not close it), and why the presigned content hash was left out — GCS
cannot bind a SHA-256 at all, and S3's binding lands as a hoisted query
parameter nothing here can verify S3 enforces.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* fix(hub): the blob seal compared two clocks that need not agree
The verification cache is sound on one argument: once a stored blob is
older than the presign TTL, no live URL for it can exist, so the bytes
cannot change again and the hash need not be recomputed.
That argument is about time, and the two times came from different
machines — o.Modified is the object store's clock, time.Since is the
hub's. A hub running ahead of storage overstates the object's age and
seals it while a minted URL is still live; a replay through that URL is
then served from cache for the rest of the process's life. NTP makes it
unlikely and a container without it, or a VM resumed from suspend, makes
it reachable.
Seal after the TTL plus an hour instead. Sealing early buys nothing —
the blob is immutable either way — so the margin costs a few extra
hashes on a young blob and removes a dependency on two clocks agreeing
that nothing in the process can verify.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72
* test(hub): age the blob-seal fixtures instead of shrinking the TTL
The clock-skew fix in
|
||
|
|
c8ab5504f7 |
feat(webapp): history can be narrowed by path, author and date (BEA-67) (#113)
* feat(webapp): history can be narrowed by path, author and date (BEA-67) The project History view was a flat scroll with no controls but per-row restore/open/download. Fine at eleven rows; unreadable after a month of agent writes, and agents write far more than people do. Four reader filters on GET /api/p/<id>/history — q= (case-insensitive substring of the path), user= (exact account), since=/until= (UTC bounds, inclusive at both ends, RFC3339 or a bare YYYY-MM-DD). They compose with each other and with the existing path=/prefix= scoping, and they are applied in the same walk as path/prefix — BEFORE the sort and the cursor skip — so next_cursor keeps meaning "the next matching entry" and paging under a filter needed no new machinery. kinds[] is still computed over every op, so a filtered view classifies edits the same as the full feed. A malformed since/until is a 400, not a silently unfiltered feed. The filter bar drives those params through the URL rather than component state: a narrowed feed is a link you can send, it survives reload, and Back undoes it. The author list accumulates across fetches — rebuilding it from the current feed would strand a reader who filtered by one author with only their own name to pick from. Zero matches gets its own empty state with a Clear. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * docs(architecture): the frontend diagram parses again The escaped quotes inside the NewProjectDialog note made the whole classDiagram fail to parse — it has been rendering as an error box, not a diagram, since that note landed. Same breakage on origin/main; noticed while adding HistoryFilters to the same file. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
a84fe444f5 |
fix(hooks): every mount gets its own hub link, so agents stop mixing projects up (#105)
A session whose root holds two connected folders got exactly one project's
base URL in its turn context: `bdrive sync --hook` emitted for the first
mount and stopped. Agents then hung the other project's paths on that URL,
producing links to project B carrying project A's path — confidently wrong,
and 404 on arrival.
The context now carries every mount as a `prefix → URL` pair, where the
prefix is the mount's path as the agent sees it from the session's folder.
A session started INSIDE a mount has no prefix to strip, so its own subpath
is baked into the base URL instead — the other half of the same bug, which
made every link from a subdirectory session miss its leading segments.
The formula also now states outright that the URL path is folder-relative
and that a path matching no listed folder is not synced and must not be
linked at all. Encoding is per-segment, `/` left literal (encodePathSegments).
Two more fell out along the way:
- `runHookSync` returns nil on every path by design ("a hook must never
fail the turn"), so the caller's `err == nil` guard could never be false.
A first mount that emitted nothing — non-hub remote, session open error —
still consumed the one emission, and NO mount got a link. It now reports
whether it produced a URL.
- stdin was read once per mount, so with several mounts only the first got
its session note stamped. The event JSON is read once per run now.
Single-mount output is unchanged apart from the folder-relative sentence.
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
||
|
|
1c703c95f8 |
feat(templates): start a project from a structure, not an empty folder (#97)
* feat(templates): start a project from a structure, not an empty folder A new project was an empty folder with a .bdriveignore in it, so every agent session invented its own layout and the folder rotted into a pile. Both surfaces now offer the same three starting points — from a template, from scratch, from an existing folder (which is just a non-empty folder, and is never restructured). internal/templates holds the shipped set as literal go:embed'ed files: `docs` (docs/, decisions/) and `para` (projects/, areas/, resources/, archives/). cmd/bdrive is one binary for the CLI and the hub, so both read the identical set — no gallery, no drift. The AGENTS.md in each is the deliverable: where a new note goes, when something is archived, what a good filename looks like. Every directory holds a real file, because BearDrive syncs paths and an empty directory would never reach a teammate. The hub seeds at creation through the existing Upload+Commit path, journaled under its own device, and records the choice on the project record — so a user who picked PARA in a browser sees PARA in the browser, and a later init cannot seed a second copy. `bdrive init --template <name>` goes through the same endpoint, with a local-seed fallback for a hub too old to know the field, and seeds in place when re-run in an already-initialized folder (the agent's post-init path). Seeding never overwrites an existing path, which is what makes a double-seed a no-op rather than a divergence. Refusals cost nothing: an unknown name and --template with --only are both rejected before any network call or write. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * test(cli): joining a project that already has a template is refused by name The one acceptance case with no test behind it: connecting to an existing project with --template must say what the project was actually created from, and must not write the other skeleton on the way out. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(templates): name the docs template in plain English, not an acronym "Plain docs + ADRs" was the recommended, first, preselected-adjacent option in a picker that non-engineers see — and it's the label people accept without reading further, so half of it not parsing is the worst place for jargon. The title also disagreed with its own blurb: "ADRs" over "docs/, decisions/", two words for the same folder one line apart. Now "Docs + decision records", which says the same thing to everyone and matches the folder names. The term itself moves into decisions/0001-record-decisions.md, where the reader is already inside the structure and the file can teach it in passing. One line in the registry drives both the web dialog and the CLI menu; the rest is prose echoing it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(templates): add the LLM wiki template The third starting point from the issue title, unblocked: the spec parked it because shipping an approximation under someone's name needed a source, and there is now one — Karpathy's LLM Wiki gist. Worth noting the issue's own one-line description of it ("few large, append-heavy topic pages") does not match the source, which is the opposite: many interlinked pages, where a single ingest touches 10-15 of them. The pattern is three layers and three operations, not a folder shape. sources/ is yours and immutable; wiki/ is the agent's and it owns every page; AGENTS.md is the schema layer — which is exactly the file this template system already treats as the deliverable, so the fit is direct. index.md and log.md ship as the two navigation files the pattern turns on. Three of the things the gist tells you to go set up, BearDrive already is: version history and collaboration (per-file history, bdrive log), an Obsidian- style reader for [[wikilinks]] (the hub viewer), and a surface for the lint pass (the dashboard is literally reads x staleness). Two rules in the AGENTS.md are load-bearing and deliberate. A page write that has not updated the index is an incomplete write — a stale index is worse than a missing page, because it is read first and believed. And with no sources yet, build nothing: the structure grows out of the material rather than ahead of it. Shipped second, not first: docs stays the recommendation because a default is the option chosen by people not reading closely, and this pattern degrades badly when half-followed. Promoting it later is one line in the registry. The shipped-template test now checks the "what happens when something stops being true" question through a set of alternatives — PARA archives, a wiki supersedes and revises — since the vocabulary honestly differs by structure. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(web): "I already have a folder" as a starting point The third way to start from the spec — adopt what you already have — had no presence in the browser. Templates and "empty" were the only visible answers, so someone with a folder of notes either hesitated or picked a template and got four directories merged into their material. The constraint that shapes it: the browser cannot reach your disk, so this cannot change what is created. It creates the same empty project "Empty project" does; what it changes is the next screen. Create therefore stays enabled — disabling it would leave the dialog a dead end AND produce no project id, which is the one thing the paste prompt actually needs. Landing on the project home with the intent, three things differ: the guide says "in the folder you already have", a note states plainly that connecting never moves, renames or overwrites anything, and the paste prompt tells the agent a folder already exists. That last one is the part that isn't cosmetic — without it an agent reads an empty project and proposes creating shared/, the one recommendation that is wrong here. It still asks which folder: that is the runbook's hard gate and nothing here weakens it. The intent rides in the URL (?connect=existing) rather than onto the project record, the same way ?v= pins a file version. It belongs to whoever is connecting right now — a teammate who connects next week has their own answer and would be told the wrong thing by a persisted flag. Five rows made the dialog tall enough to push Create off a short viewport, so .modal scrolls internally. A hairline divider between the seeding and non-seeding rows was tried and removed: --border is 7% white, which at 1px in a gap renders as literally nothing. The gap is the cue that reads. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * feat(web): with no projects, open the create dialog and give the page a way in A signed-in account with no projects landed on a page whose only path forward was pasting a prompt into a coding agent. Now the create dialog opens itself — with nothing to browse there is nothing else on that page to do — and the page behind it leads with "Start a project" and a button, so closing the dialog is not a dead end. The dialog moves up to HubApp because three things ask for it now: the sidebar's +, the empty state's button, and the auto-open. ProjectNav keeps only an onNew callback; one owner beats three copies of the create handler. Two guards on the auto-open. It fires once per mount, keyed off a ref rather than the empty state, or closing it would immediately reopen it. And it never fires on a read-only hub, which refuses creation server-side with a 403 — opening a dialog that cannot succeed is worse than the page it covers. The agent paste-prompt stays, demoted to "Or let your agent do it": it is still the right path for someone who wants the folder connected in the same breath, and it is the only path on a hub where this account cannot create. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
b408e004b3 |
fix(history): stop reporting the device IP on every change (BEA-43) (#81)
* fix(history): stop reporting the device IP on every change (BEA-43) The history API embedded the whole DeviceInfo, so every project member read a teammate's server-observed IP (plus user/last_seen) next to every change on a page whose job is "who changed this file". Project a three-field historyDevice instead — id/name/os — mirroring heatByDevice. The device registry is unchanged: Observe/requestIP and both MetaStore backends keep recording the IP. This is a response projection, not a change in what gets collected. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * docs: history shows device name and OS, not the IP (BEA-43) README, SKILL.md and the docs site all promised the History view would show the connecting IP. It no longer does — the registry still records it. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
4dfe6f44f0 |
feat(history): undo a file an agent run created (BEA-35) (#82)
History could restore an edit or a deletion, but a file a run CREATED was the one thing it couldn't reverse — the ADDED row said so in copy and offered no button. The missing capability was a hub-written delete op: restore.go only ever journaled puts. POST /api/p/<id>/remove journals exactly one journal.KindDelete op under the hub's own device identity, behind restore's gates (gateUpload, PermWrite, cleanUploadPath, quota CheckWrite/RecordUsage) plus a volume- snapshot existence check so the API 404s on what the tree doesn't show. Commit's journal-append tail moves into RemoteSource.appendOp, which both writes now share — one writer per journal, unchanged. The ADDED-in-a-run row gets an "undo — remove file" control that confirms first (it reaches every synced device), and the DELETED row it leaves behind restores the file with its original bytes. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
872ba702cb |
fix(log): bdrive log reads as a timeline — newest first by edit time (#BEA-40) (#90)
`bdrive log` sorted by the lamport clock, so the wall-clock stamps it prints came out non-monotonic — two 06:09:24 rows above a 06:10:00 row. And every op of one scan is stamped with that scan's commit time, so a 22-file agent run collapsed onto a single stamp. Neither is readable as a timeline, which is the whole job of the command. Two display-only changes: - `journal.Op` gains `Mtime` (`omitzero`, so old journals and old binaries are unaffected), populated on put ops from the `os.FileInfo` the scan already holds. Deletes and conflict copies keep their commit time. - `syncer.DisplayTime` / `SortForDisplay` order by the timestamp that is actually printed, ties broken by reversed `journal.Less`. `bdrive log` sorts and *then* truncates, so `-n 25` is the 25 newest by that stamp. `journal.Less`, `Sort`, and `Replay` are untouched — replay order is the convergence contract, so the sort lives in `syncer`, not in `journal`. `LogEntries` also keeps returning causal order because `bdrive restore` walks it to find a file's previous version. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
b1c0bba415 |
feat(auth): the browser sign-in says whose account the terminal gets (#96)
* feat(auth): ask before signing a terminal in as whoever the browser is `bdrive login` opened /auth/cli and the browser bounced straight back with a code. Whoever the browser happened to be signed in as is who the terminal became — silently. That is frequently not the account the user meant: a personal login left open in the default browser, a teammate's session on a shared machine. The mistake surfaces much later, as a synced folder full of commits authored by the wrong person, which is far more work to undo than one click would have been. The device flow already got this right in #83 — it names the account, offers to switch, and says what approving grants. The browser flow said nothing at all, for the same outcome: a token that acts as you. So /auth/cli now confirms first. GET renders the page (who you would be signing in as, a Switch account link that comes back to this same pending sign-in, what is asking, and where it is waiting); POST is what mints the code and redirects to the loopback listener. A GET therefore grants nothing, so a link someone else got you to open can no longer mint a code on your behalf. whoBlock loses its pendingGrant parameter and renders only the identity half. What is asking differs per flow — a device has a name and an OS, a CLI on this computer has a loopback port — so each page now renders its own rows through a small helper instead of whoBlock pretending to a shape neither quite fits. The CLI's own wording follows: "waiting for you to approve the sign-in in your browser", since being signed in already is no longer the whole story. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72 * docs: the browser sign-in confirms first, and says whose account it grants README, the CLI reference, and the self-hosting auth page all described the old behaviour — sign in and the page bounces a code straight to the terminal. They also read as though only `--device` had an approval step. Both flows now confirm; say so, and say why it matters (the browser session is often not the account the user meant). Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72 * feat(auth): one web step for a first sign-in, not two The confirmation page fixed the wrong-account problem and created a smaller one: a user with no browser session now saw two pages on their first `bdrive init` — sign in, then approve — where the sign-in had already settled the only question the second page asks. So authenticating *for* a pending CLI sign-in now counts as approving it. The login and signup pages carry a line saying a terminal is waiting and that the account used here is the one it will act as, which is where that consent is made informed; reaching the callback then needs no second click. The marker is server-side, bound to the exact pending sign-in, single use, and two minutes long, so it can only ever skip the page it was granted for and only once. It cannot be forged: setting it requires authenticating as that account, and anyone who could do that could click Approve anyway. An existing session still gets the page — that is the case where the browser may be signed in as someone the user did not intend, which is the whole reason it exists. Net effect: exactly one web interaction either way. The device flow keeps its explicit approval. Its page names a machine that isn't this one, along with the OS and address it came from — information no login form can convey, about a grant to somewhere else. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72 * fix(auth): keep `bdrive login` on one line in the approval hint It wrapped mid-phrase into two separate code boxes, which reads as two commands rather than one. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72 * refactor(auth): one approval page for both sign-in flows The two flows ask the same question — shall this thing act as you? — and had two copies of the page asking it, differing in three strings. They had already started drifting: a wrapping fix went into the CLI copy only, leaving the device page able to break `bdrive login --device` across two code boxes. A page whose whole purpose is consistent disclosure is a bad place to keep two of everything. So pageAuth owns the shape (session check, redirect to login, whoBlock, rows, the Approve form, the note) and each flow supplies an authRequest describing what differs: how the request is identified, what is asking, and what approving does. Two asymmetries are now explicit rather than accidental. freshAuthSkips is true only for the local flow — signing in and approving are the same act when the terminal is on this machine, and are not when the token goes to another one. live() reports whether the request still exists, because the device flow's link expires while the CLI flow carries its whole request in the URL and has nothing to expire. detail is a function, not a slice: the device rows come off the pending grant, which only exists after live() has found it. No test changed. The pages render byte-identically — same sha256 for all three CLI screenshots before and after — and the device flow was driven end to end against a real hub, approving a real `bdrive login --device`. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72 * feat(auth): both sign-in flows always ask you to approve Consistency between the two flows is worth more than the click it saves. Letting a sign-in count as its own approval made the local flow one step and the device flow two, so the same product asked for consent in two different shapes depending on which machine you were on — and the shape that skipped it was the one where the page had something to tell you. So the fresh-auth marker is gone: sign in, then approve, on both flows. That drops a map, two methods, a descriptor field, and a branch in pageAuth — the unified handler now has exactly one path through it. A first `bdrive init` on a fresh machine is two web pages again. That is the deliberate trade: the approval page is where a user sees which account a machine is about to act as, and nothing shortcuts it. The sign-in page keeps the line saying a terminal is waiting. It no longer carries the consent — the next page does — so it is there to explain why a password prompt appeared at all. TestBothFlowsAlwaysAskToApprove replaces the one-step test and runs the same assertions over both flows as subtests: no session sends you to sign in carrying the request, signing in returns to the request without granting, the approval page is there every time, and only the POST grants. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DgF8JsoeNPVShGYWdooE72 --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Snow Lee (Sungwon)andGitHub
|
118abc67e2 | feat(cli): "bdrive serve" replaces "bdrive web" (web stays as an alias), and the README leads with the agent install (#94) | ||
|
|
fd4f5c7964 |
feat(onboarding): folder name follows the project name, and project ids are UUIDs (#89)
The paste prompt now carries the project's name so an agent recommends a folder of that name; with no project at all the recommendation is `shared/` (and `bdrive init shared` names the new project after the folder), replacing the old `wiki/` default. New project ids are UUIDs instead of `p-` + 8 hex chars. The route validator still accepts the legacy shape — ids are permanent — and the client-side URL parsers (remote/http.go, bdrive share) now only check the shape of a URL segment, leaving the hub as the single authority on which ids are valid. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
fb6ce347c4 |
feat(daemon): survive a reboot — login autostart on macOS/Linux/Windows, and a lock instead of a pidfile (#88)
* feat(daemon): bring sync back after a reboot, and stop trusting the pidfile A reboot killed every daemon and nothing restarted them. Agent hooks still synced per turn, which is what made it easy to miss: a folder looked fine while an agent worked in it and went stale the moment one didn't. `bdrive init` now registers a login item (macOS: a user LaunchAgent) that runs the new `bdrive resume` — one registration per machine, which starts a daemon for every enrolled, unpaused mount, so adding a project later needs no re-registration and `bdrive stop` still means stay stopped. `--no-autostart` opts out, `bdrive autostart install|uninstall` manages it. Writing the plist is the whole job: no `launchctl` shell-out. launchd loads agents at login anyway, the caller has just started the daemon for this session, and shelling out would let a test or a packaging script register a real login item as a side effect. The recovery path was also broken, which is why this is one change. Liveness came from `kill(pid, 0)` on daemon.pid — but that file lives in $BDRIVE_HOME and survives the reboot that killed its process, so any same-user process recycling the pid read as a live daemon. `bdrive status` said "running", and worse `daemon.Start` returned early, so the one documented recovery (`bdrive init`) reported success and started nothing. Liveness is now an flock held for the daemon's lifetime: the kernel drops it at death or reboot, and it makes two daemons on one mount impossible. The pid stays for display and for signalling. internal/autostart is darwin-only today; autostart_other.go returns ErrUnsupported and every caller already treats that as "nothing to do", so Linux (systemd user unit) and Windows are one file each. Tests: internal/daemon gets its first ones — a recycled pid must not read as running (the exact regression), the lock decides liveness, a second holder is refused. internal/autostart covers write/idempotency/stale-path-rewrite/ uninstall with HOME redirected, and lints the plist with plutil so launchd can actually parse it. The CLI e2e asserts init registers the agent, that it runs `resume`, that resume finds the live daemon instead of starting a second, and that --no-autostart is silent. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016aYntCWwdUhpzUfEk3ddyJ * feat(autostart): Linux support — a systemd user unit alongside the launchd agent Same three functions, same discipline. Linux writes $XDG_CONFIG_HOME/systemd/user/beardrive.service (Type=oneshot, no Restart= — `bdrive resume` exits by design) plus the default.target.wants symlink that `systemctl --user enable` would create, because systemd ignores a unit nothing wants. No `systemctl` shell-out, for the same reasons as launchctl: the file is the registration, it only matters at the next login, and a container or ssh session has no session bus to talk to. Install declines with ErrUnsupported unless systemd is actually the init system (/run/systemd/system, i.e. sd_booted) — on Alpine, WSL1 or a slim container a unit file is inert decoration, and reporting "registered" would be a lie. Installed() likewise requires the enable symlink, not just the unit: a unit nothing wants never starts. os.UserConfigDir honors XDG_CONFIG_HOME, so relocated config dirs work. Windows is now the only gap; autostart_other.go is !darwin && !linux, and the shared writeIfDifferent/selfPath moved into the tag-free autostart.go (darwin now uses them too). Tests run on Linux, not just compiled for it: cross-compiled test binaries executed in a container, both with /run/systemd/system present (unit written, enabled, idempotent, stale ExecStart rewritten, broken symlink repaired, XDG honored, uninstall removes both) and without it (Install declines and writes nothing). The daemon flock tests were run there too, since flock semantics are per-OS. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016aYntCWwdUhpzUfEk3ddyJ * feat(autostart): Windows support — a per-user Run entry Third platform, same three functions. Windows has no user service manager in the launchd/systemd sense, so the registration is a HKCU\...\Run value via golang.org/x/sys/windows/registry (already in the module graph; go mod tidy just promotes it to direct). Chosen over the alternatives for the same reason the other two write files: no admin rights, no COM (a Startup-folder .lnk needs it), no schtasks shell-out. It is also honestly discoverable — the entry appears in Task Manager's Startup tab, where someone can disable it without knowing bdrive exists. The executable is quoted because Explorer parses the value as a command line and Program Files has a space in it. Two things a reader should not have to discover for themselves: - The tests here have NEVER RUN. They are written and compile-checked (GOOS=windows go test -c) from macOS; there is no Windows host or usable container on an arm64 mac. They execute the first time the suite runs on Windows. They also cannot use a temp HOME the way the macOS and Linux tests do — HKCU is real — so each one snapshots and restores the previous value. - `GOOS=windows go build ./...` still does not pass, and this package is not why: internal/store's Lock uses syscall.Flock and internal/daemon uses syscall.Kill and Setsid, all unix-only (true before this branch too). A Windows port means LockFileEx plus a stop story for a platform with no SIGTERM — a separate change, against the sync invariants, and untestable from here. So this code is correct and currently unreachable. autostart_other.go is now !darwin && !linux && !windows (the BSDs). Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016aYntCWwdUhpzUfEk3ddyJ --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
31f705e287 |
chore: drop the Claude plugin and the bundled skill — hooks are the integration (#85)
Two front doors for the same setup, and one of them was a second copy of everything. The plugin shipped the skill the CLI already installs (to four platforms, not one), hooks that ran the identical commands `bdrive hooks install` writes machine-level, and an install flow duplicating INSTALL_FOR_AGENTS.md. Nothing deduped, so a machine with both pulled twice per turn — two blocking syncs — and showed two identical `beardrive` skills in the picker. What remains is `internal/agenthooks` plus the runbook: init registers a blocking pull (which also injects the gated-link convention as additionalContext), an async push on Write/Edit, and read-log for the heatmap, in each platform's user config, once per machine. That is the whole integration, and it is the part that was never optional. Removed: plugin/, .claude-plugin/marketplace.json, internal/agentskills, `bdrive skill`, `bdrive hook-approve` (its PreToolUse auto-approve only ever helped when a plugin pre-installed it; the substitute is a `Bash(bdrive:*)` permission entry, which is user-owned config and needs no code). The e2e now asserts the absence: no SKILL.md in any platform's skills dir after init, and no `skill` subcommand. login_test keeps the "no revoke surface" wording check on logoutNote alone. Claude-Session: https://claude.ai/code/session_016aYntCWwdUhpzUfEk3ddyJ Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
dfb9da3260 |
feat(auth): one approval link for device sign-in, and a page that says what it grants (#83)
The headless flow printed a short code to retype into a bare "Approve" form. Now `bdrive login --device` prints a single link — the token lives in the path (/auth/device/<token>), so there is nothing to read off one screen and type into another. The page it opens is a consent page rather than a text field: it names the account the device would act as, offers Switch account (logout now honors ?next, so you land back here), and shows the device name, OS, and the address the server observed. That matters because this flow's weakness is a stranger sending you their pending link; an anonymous "Approve" gives you nothing to notice with. Approval is still a POST from the page, so a link alone cannot grant, and SameSite=Lax keeps a cross-site form out. Also aligns the /auth/* pages with the app's tokens, which had drifted: card #0c0e10 vs --color-card #15171b, 8px controls vs --radius-ctl 7px, hand-picked #ff9b91/#6fd699 vs --color-del/--color-add. The style block now declares the tw.css tokens by name and every rule uses them. Older CLIs still print /auth/device?code=…, so that shape 303s to the path form; a pre-0.13 hub returning no verify_url still gets the old type-the-code instruction from the CLI. Claude-Session: https://claude.ai/code/session_016aYntCWwdUhpzUfEk3ddyJ Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
4ff92c56ab |
feat(history): group agent runs and restore any version (BEA-6) (#69)
* feat(history): group agent runs and restore any version (BEA-6) BearDrive recorded everything and could restore nothing. Now every version of a file has a Restore button — in the hub's History view and as `bdrive restore` — and the changes one agent run made read as one card instead of N loose rows. Restore is a NEW put op pointing at the old blob: journals are never rewritten, so one-writer-per-journal holds and peers converge on the restore like any other edit. The hub reuses RemoteSource.Commit (the upload commit minus the upload); the CLI writes the bytes into the working folder and lets the ordinary cycle journal them, so the sync engine gains no new write path. Grouping is a pure frontend group-by on (note, device) over the existing /history response — no journal or API change. Known gap, stated in the UI and the docs: nothing in the hub writes a delete op yet, so a file a run *created* cannot be un-created. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * fix(history): don't repeat a run's note on every row in its card UI pass on the real hub: inside a run card the note is the card's header, so printing it again on each row said the same thing N times. The header now carries the note (linkified, so an agent's session link still opens) and the collapse control is its own button rather than the whole header — the link could not live inside a button. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
dcd0517e92 |
feat(cli): bdrive scope --explain — prove what leaves this machine (BEA-24) (#70)
The local-first claim was asserted, never demonstrated: nothing anywhere told you what your laptop chose *not* to send. `bdrive scope --explain` walks the folder and prints two sorted lists — synced and not synced — with counts and a pointer at what it does not answer. The decisions come from the same walk the sync cycle uses. scan()'s WalkDir decision tree moves into walkFolder (internal/syncer/walk.go), the only copy of the rules; scan and Explain both go through it, so the output provably cannot drift from real sync behavior. Pure read: its own Filter, no Session, no volume flock, no network. Fully-excluded directories collapse to one counted line; nested mounts are annotated as syncing through their own project rather than called "not synced", which would be a lie in a trust surface. Known gap, deliberate: this answers "what leaves from now on", not "what is already on the hub" — the footer points at `bdrive forget`. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
acb99e85e3 |
feat(hub): set an expiry on a share link from the UI (BEA-29) (#74)
The CLI has had --expires all along and the Public links table has always
said "no expiry" — naming an alternative the hub UI never offered. Every
link minted from the browser was permanent.
Expiry is offered AFTER minting, not before: the one-click share stays one
click, and PATCHing the token we just handed out keeps the URL already on
the clipboard valid. Minting with a TTL would instead create a second link,
since ShareDB.Create only reuses permanent ones.
- ShareDB.SetExpiry re-dates a live share in place; repo-write failure
restores the previous row rather than deleting it (unlike Create's
rollback, this row already existed — dropping it would revoke a live
link over a disk hiccup).
- PATCH /api/shares/{token} mirrors handleShareRevoke: resolve the token,
requirePerm(PermWrite), act. Duration parsing is copied from
handleShareCreate so the two routes can't drift.
- Share dialog gets an Expires select (Never / 24h / 7d / 30d). A failed
PATCH toasts and reverts the control. Copy link keeps the dialog's
initial focus — the new control would otherwise have taken it.
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
|
||
|
|
4e34d03e14 |
feat(hooks): user-scope agent sync hooks, one-command setup, --only scoping (#71)
* feat(hooks): register agent sync hooks per machine, not per project Agent platforms read hook config only from the directory a session starts in — never a parent, never a subfolder. Project-level hooks therefore fired only for sessions that happened to start at the mount, and, living inside a synced folder, they replicated one machine's agent config to the whole team (a second writer of a file bdrive already owns). Claude Code additionally ignores project hooks until the folder is trusted, so in practice they were often inert without any visible sign. Hooks now go to each platform's user config, once per machine, covering every session in every folder; the existing shell guard keeps them a no-op outside BearDrive projects. Install migrates away blocks older versions wrote into projects, and `bdrive hooks uninstall` removes ours while leaving foreign hooks untouched. Setup is also one command now. init absorbs the skill install, prints the hub link, and takes --server, so connecting to a named hub no longer needs a separate login; the runbook forbids preflight and command chaining, since each distinct command costs the user a permission prompt. For plugin users a PreToolUse hook auto-approves bdrive's own setup subcommands — narrowly: any shell operator in the command disqualifies it. Also drops --shared in favor of `init . --only wiki,docs`, which writes a managed block of .bdriveignore rules instead of a second scope mechanism. Because those rules sync, `sync --prune` now refuses on a scoped project rather than stripping everything outside the scope from the hub for everyone. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016aYntCWwdUhpzUfEk3ddyJ * docs: fix stale claims an audit found against the new CLI An audit of every doc surface against the code turned up claims that the user-scope hook move and the one-command init made false: project-level hooks "riding the repo", the Claude trust prompt, Codex's //hooks project layer, `--no-hooks` skipping the skill (it does not), prune reconciling against a per-device scope (it now refuses on a scoped project), and `--scan-interval`/`--remote-interval` documented as init flags when they only exist on `bdrive daemon run`. Also documents the surface added today — `--server`, `bdrive hooks uninstall`, and the plugin's PreToolUse auto-approval — refreshes the two sample `init` transcripts to the real output, and corrects hook matchers that had drifted from agenthooks.go. `bdrive scope` told users to narrow an existing mount with `bdrive init . --only <dirs>`, which resume then ignored — a dead end. Init now applies --only on resume, writing the scope block, so the advice works. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016aYntCWwdUhpzUfEk3ddyJ --------- Co-authored-by: Claude Opus 5 <noreply@anthropic.com> |
||
|
|
9088127176 |
feat(sync): bdrive forget + sync --prune to take ignored paths off the hub (BEA-20) (#68)
Adding a path to .bdriveignore only stopped future uploads: anything that synced before the rule existed stayed on the hub forever, with no command that removed it without deleting it from local disk on every device. Two engine changes make an explicit removal safe: - materialize's delete loop now consults the filter. A cached path absent from the replayed target that the rules exclude is dropped from tracking instead of unlinked — without this, any delete op for a now-filtered path wipes every peer's local copy, which is the data loss this issue is about. - the filter is reloaded mid-cycle from the pulled .bdriveignore, before materialize. A peer receiving the new rules and the deletes they justify in one batch would otherwise materialize with stale rules and the guard would never fire. materialize's write side is split into materializeFile so the ignore file can land on its own. On top of that, Session.Prune journals a delete for every path the replayed state still holds that the SHARED rules exclude — reconciling against the replay, not the local cache, because a path filtered out in an earlier cycle was dropped from the cache back then and is invisible locally today. The rules are deliberately ignore-only: the include scope lives in each device's own .bdrive/config.json and does not sync, so pruning against it would let a narrow-scope device delete a whole-folder teammate's files. Plain `bdrive sync` and the daemon are unchanged — pruning is never a side effect of editing .bdriveignore. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
589be3a6de |
fix(hub): open the project Dashboard to every member, rename /insights → /dashboard (BEA-12) (#62)
* fix(hub): open the project Dashboard to every member, rename /insights → /dashboard (BEA-12) "Dashboard" was the first sidebar item a new member clicked and it always refused: it landed on /<project>/insights showing "Insights is for hub admins and org owners." The gate was client-side only — GET /heat is gated on project membership and returns counts without actor identities, so every member's browser could already fetch every number the page draws. Drops the canInsights gate (nav item, dedicated route, project-home embed, ⋯ menu entry) and renames the view route insights → dashboard so the nav label, the URL and the page title finally agree. The shipped /insights URL still resolves and normalizes to /dashboard (LEGACY_VIEWS in router.ts) so bookmarks don't 404 and only one URL stays live. No server change: /heat gating and response shape are untouched. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> * docs(architecture): router VIEW_ROUTES now names dashboard, with LEGACY_VIEWS Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
Snow W. Lee (Sungwon)andGitHub
|
65d4df5a84 | One paste-able URL onboards any agent: INSTALL_FOR_AGENTS.md (#57) | ||
|
|
056c883204 |
fix(sync): anchor --shared include entries to the mount root (BEA-5) (#56)
`bdrive init --shared wiki` wrote `include: ["wiki/"]`, which compile() treats as an unanchored gitignore pattern — so any nested directory named `wiki` synced too. Shared-subfolder mode is what people use to keep private material out of a project, and it was silently widening the scope: 15 files under .agents/, .claude/ and .gemini/ leaked into a real project from .../detector/shared/ dirs. cleanShared now emits "/wiki/", which fixes both callers (init --shared and bdrive scope add). config.LoadProject anchors legacy single-segment entries on read, so the existing mounts are fixed without a re-init — and that also keeps `bdrive scope rm wiki` working against pre-fix configs, with a belt-and-braces unanchored candidate key in scopeRemove for any config that bypasses LoadProject. Not touched: compile() itself, and no delete op for the already-leaked remote files (BEA-20 — a delete would unlink teammates' local copies). Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
2c85747464 |
feat(sync): .bdriveignore always syncs, regardless of scope or rules (#54)
The ignore file is exempt from filtering in Filter.Skip: it syncs even on --shared mounts (where it sat outside the include list and was local-only) and even when one of its own patterns matches it. One guard covers both scan and materialize since they share the filter. Docs updated on all three surfaces (SKILL.md, install.md, web/docs). Claude-Session: https://claude.ai/code/session_01G3AdFdps7seYbW6FhyK58t Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
0236e1b272 |
feat(cli): multi-folder --shared at init + bdrive scope to edit the sync scope (#53)
* feat(cli): init --shared accepts multiple subfolders (repeatable or comma-separated) --shared is now a slice flag: `--shared wiki --shared docs` or `--shared wiki,docs` sync several subfolders into one project (include list ["wiki/", "docs/"]). The interactive scope prompt takes a space- or comma-separated list. Entries resolving to ".", "", or ".." error out — silently dropping them would widen scope to the whole folder. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01AubcaQscjcQucXwh578vqG * docs(plugin): skills/commands propose multiple --shared folders at init The init/install flows now scan for all knowledge folder candidates and offer them as one --shared list (one project, one permission set), noting that folders needing different access belong in separate projects. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01AubcaQscjcQucXwh578vqG * feat(cli): bdrive scope — add/remove shared subfolders without editing JSON `bdrive scope` shows the include list; `scope add`/`scope rm` edit it from the mount root. The daemon re-reads config each tick, so changes apply in seconds. rm deletes nothing (newly filtered paths drop from the cache with no delete op); removing the last entry is refused since an empty include list means whole-folder sync. add onto a whole-folder project is refused for the same narrowing hazard. Skill/README/docs updated; scope added to the cli-sync diagram's command list. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01AubcaQscjcQucXwh578vqG * docs+cli: scoping guide covers multi-folder --shared and bdrive scope; init hints on ignored --shared at resume The scoping guide (the dedicated page for this feature) now shows --shared wiki,docs and a "Change the scope later" section for bdrive scope; setup-by-hand and project-files point at it. init resume with an explicit --shared now says the flag is ignored instead of staying silent. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01AubcaQscjcQucXwh578vqG --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
69e7231a70 |
feat(hub): per-project permissions — none/read/write/admin, invite-only projects, honest degraded sync (#46)
Access was binary and org-wide: any org member got full read+write on every project. Now each project carries four ordered levels, resolved by one resolver and enforced at one choke point. - `projectPerm` (perms.go) replaces `projectAllowed`; `proj(level, h)` in server.go gates every per-project route by the level it declares at registration, so no handler grows its own check. - `Project` gains Creator/Default/Perms. `Default == ""` means write, so an upgraded hub behaves identically until someone edits permissions. - Creator becomes the first project admin; org owners are implicitly admin everywhere in their org and a grant naming one is refused, not ignored; a project always keeps at least one explicit admin. - Default `none` makes a project invite-only. A `none` member is treated exactly like a non-member, including on create-or-join by name. - Rename/delete move from org-owner-only to project `admin`. - Both metadata backends persist it: the file store rides along, the SQL store gains `project_perms` plus an idempotent ALTER for the two new columns (migrate() had only ever created tables). Client side, a refusal stops looking like an outage: `remote.ErrForbidden` plus `Result.ReadOnly` (push refused → pull-only) and `Result.NoAccess` (pull refused → paused, working folder untouched). Neither sets Offline, neither loses a local op, and re-granting self-heals on the next cycle. `bdrive status`/`sync` and the daemon (once, on transition) say which. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
|
6628718f1d |
fix(cli): agent hooks never sync folders this device didn't opt into (#44)
* fix(cli): agent hooks never sync folders this device didn't opt into The turn hooks decided "this folder is managed" from the mere presence of .bdrive/config.json — a file designed to travel with the folder. Two holes: - A config.json arriving via git clone / copied dir made one hook firing silently mint a device identity, register the mount, create a volume store, journal the whole folder, and inject the hub-link formula — on a device that never ran init or login. - `bdrive stop` only killed the daemon: the next agent turn's `bdrive sync --hook` resumed a full sync cycle and kept injecting links, and `stop --forget` was undone within one turn by registry self-heal. Fix: one gate (`syncBlocked`) in the paths all hooks route through — sync/sync --hook/read-log now require the mount to already be enrolled in this device's mounts.json (read without ResolveMount's enrolling self-heal) and not paused. Hook mode exits silently; plain `bdrive sync` errors with a `bdrive init` pointer. New per-device paused marker in the volume dir: set by `bdrive stop`, cleared by `bdrive init` (startSync). Only init enrolls or resumes; folder moves still self-heal since enrollment is keyed by mount id, not path. Docs updated (README, SKILL.md, docs cli reference, CHANGELOG). Tests: hook/read-log no-op + no-enrollment on unenrolled and paused mounts, plain-sync refusals, stop→pause→forget regression, paused marker contract. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: architecture-diagram PRs must show before/after excerpts of changed classes The "Architecture changes" PR section now names exactly what changed and shows Before and After mermaid excerpts scoped to the affected classes and their immediate relationships — never the full diagram (Before = merge base). Convention updated in CLAUDE.md, architecture/README.md, and the pre-PR hook's reminder text. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: add cli-sync architecture diagram; widen diagram convention to the CLI architecture/cli-sync.md draws the CLI and sync engine (cmd/bdrive + internal/{syncer,store,journal,config,daemon,agenthooks}): the Session cycle over Store/journal/remote, and the command layer with the new syncBlocked opt-in gate, paused marker, and enrollment ownership. The pre-PR hook and CLAUDE.md now watch these packages too, so CLI-side structural changes trigger the before/after-excerpt convention the same way server changes do. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: full-coverage architecture diagrams — overview, frontend, agentskills Every application package is now drawn somewhere: overview.md (system diagram — package map, device↔hub↔storage flow, agent surfaces, and the private cloud/ repo as an external seam consumer), webapp-frontend.md (the hub SPA's modules: App/HubApp/VolumeApp/Browser, the in-repo nav/router, api layer, hooks, components), and agentskills added to cli-sync.md. The pre-PR hook now watches all of cmd/, internal Go code, and frontend/src (generated static/ excluded); CLAUDE.md and architecture/README.md state the coverage rule: every code change lands in exactly one detail diagram's scope, web/docs and cloud/ deliberately excluded. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs: PR bodies start with a TL;DR — max 5 informal one-liners Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
4f5f517a6f |
feat(cli): bdrive export/import — move projects between hubs with full history (#43)
The anti-lock-in story for cloud-hesitant users: export a project's complete store (every device's journal + every blob, i.e. full history and authorship) into a portable tar.gz, and import it as a fresh project on any other hub — cloud → self-hosted or back. The archive is simply the remote store layout plus a manifest, streamed through the existing remote.Backend, so no server-side support is needed and it works against every existing hub. Import verifies each blob's content hash, rejects foreign tar entries, requires an empty target project, and refuses journal-less archives. Reconnecting devices resume exactly where they were, because their journals are byte-identical. Docs: README + SKILL.md command tables, docs-site CLI reference section, and a new step-by-step reference page (Migrate between hubs). Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
220e27a9c2 |
review fixes: gate init next-steps on background mode, shared stdinIsTTY, daemon reconnects on token change, whoami surfaces settings errors, doc staleness (#41)
- init -f no longer prints 'daemon now keeps this folder in sync' after the foreground daemon has exited - one stdinIsTTY() helper (TTY or Cygwin pty) shared by init's prompt gate and login's headless fallback — the two sites disagreed on Cygwin - the daemon drops its remote backend when the device token changes, so an account switch mid-run reconnects with the new credential instead of pushing with the old one (httpBackend captures the token at open) - whoami reports a settings read error instead of claiming 'not signed in' - self-hosting/authentication and manual/setup-by-hand now describe the automatic device-code fallback instead of presenting --device as required Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
|
34afef17cc |
fix(cli): onboarding friction — headless login fallback, --version, init next steps, labeled authorship (#39)
BEA-7 launch-critical set from the onboarding audit: - login: shells without a TTY auto-fall back to the device-code flow (agents/CI/SSH no longer hang on the browser callback); a failed browser open also falls back, and the waiting state hints --device - bdrive --version now works (cobra root Version), same output as bdrive version - init prints a next-steps block: daemon auto-sync note, the 'bdrive init --project p-xxx' connect command for teammates, bdrive log / bdrive share - authorship: the daemon re-reads settings.json every tick so a login/logout/account switch is reflected in op authorship instead of stamping a stale identity forever; whoami now shows the signed-in account and labels the git/OS author as the signed-out fallback Docs updated in README, plugin SKILL.md, and web/docs reference/cli.md. Co-authored-by: Claude Fable 5 <noreply@anthropic.com> |
||
|
Snow LeeandClaude Opus 4.8
|
16210f3236 |
docs: use case — give coding agents the business context
Add a job-shaped use-case page for engineering teams whose customer context lives where coding agents can't see it: sync a context/ folder into the repo via --shared, gitignore it (one-writer invariant), point AGENTS.md at it, and run the in→used→back loop. Register it in the sidebar between team-wiki and company-brain. Also sharpen team-artifacts' "send a person a link" story — markdown/HTML render as pages at a URL, member-only vs public links. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_012sSVdMviADW8pu6sAJ4SfX |
||
|
Snow LeeandClaude Opus 4.8
|
93f2c6bc84 |
docs: move Use cases below Manual setup
Sidebar order is now Start here -> Working with agents -> Manual setup (optional) -> Use cases -> Self-hosting -> Reference -> Concepts. README and CLAUDE.md carry the group order and the rule for what belongs in each, so both move with it. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
Snow LeeandClaude Opus 4.8
|
16ababe37d |
docs: use cases — five job-shaped pages that route into the guides
"Is this for me?" was answerable only by reading the guides and doing the translation yourself. These five pages answer it directly, sit between Start here and Working with agents, and route out rather than re-teaching features: - Share work across your team's agents — a team that doesn't live in a terminal: Cowork and Claude Code share plugins, so the agent does the setup and nobody opens a shell. - Keep a wiki your agents maintain — knowledge written as a side effect of work; Insights' hot-and-stale quadrant is the maintenance queue. - Turn a personal brain into a company brain — OKF bundles, gbrain repos, Obsidian vaults. They are already markdown directories, so there is nothing to convert. - Run a personal wiki, publish part of it — history as the point, plus per-file public links. - Carry one context across agents and devices — one project, many mounts, and what actually happens when two machines edit one file. Titles are job-shaped; the persona is named in the first line and in the description, which is also the search snippet and the llms.txt line. The company-brain page is the long one (790 words vs ~400) because it carries two frictions worth being honest about. gbrain's own team setup shares a brain through a remote Postgres, an HTTP MCP server, and per-teammate OAuth with isolation enforced in SQL; file sync plus each person's local brain skips all of that at small scale, and the page says what the server still buys you rather than dunking. And privacy does not map cleanly: gbrain scopes per person, BearDrive's unit of membership is the project, so a walled boundary is a separate project — stated plainly, with a table. Unverified: whether "let one machine run consolidation" matches how gbrain teams actually work. Written from the docs, not from practice. Verified: 23 pages build, zero broken internal links, both external references (Google Cloud's OKF announcement, the gbrain repo) resolve 200. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
Snow LeeandClaude Opus 4.8
|
8f239dfae4 |
docs: reorganize around the agent path, CLI becomes optional
The sidebar and the homepage disagreed. index.md's "Where to start" already led with connecting an agent, but the left rail read Install (brew) -> Quickstart (bdrive login, bdrive init) -> ... -> Connect an agent, three groups down. Anyone following the rail met the CLI first and the skill last — the opposite of how the product is meant to be adopted. The agent page was also filed under guides/, which this repo defines as agent-workflow docs rather than setup. Sidebar order is now the recommended path, and that path is agent-first: Start here what it is -> set up with your agent -> your first hour Working with agents shared memory, artifacts, read heat, scoping Manual setup (opt) install the CLI, set up by hand, skills and hooks Self-hosting / Reference / Concepts unchanged - start/setup (was guides/connect-an-agent): rewritten as the front door. Claude Code's plugin, then the one-paste for Codex/Gemini/Hermes, then what the agent just installed and how to check it. - start/first-hour (new): the page that was missing — ask for a doc, get a link back, share it, a teammate's agent picks it up. What success looks like without a command you have to type. - manual/skills-and-hooks (new): the mechanics lifted out of the old onboarding page — per-platform paths, hook events, idempotency, project-level vs per-user — so the Start page can stay conversational. - manual/install and manual/setup-by-hand (were start/*): both now open by saying you probably don't need them, and link back to the agent path. - index.md leads with "You don't install it — you ask your agent to"; the CLI and hub sentence moves below it. No `brew install` appears anywhere in Start here. Reference -> CLI stays exactly where it was: the people most likely to self-host are CLI-first, and burying it would read as condescending. Three public URLs moved, so astro.config.mjs declares redirects. Static builds emit meta-refresh only, so README carries copy-paste 301 rules for the host — Firebase Hosting and a Cloud Storage + load balancer URL map. Verified: 18 pages build, zero broken internal links across the built output, all three redirects resolve. CLAUDE.md and the docs README record the rule so this doesn't quietly revert. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
Snow LeeandClaude Opus 4.8
|
25f890c03f |
feat(brand): the Stack mark and Jersey 10 across app, auth, and docs
Replaces the 🐻 emoji standing in for a logo everywhere. The mark is the letter B built from three rectangles — a rail and two blocks, the same shape as the product (a spine with volumes hanging off it). One fill, so `currentColor` themes it in the sidebar, the favicon, and flat ink. - Web app: <Mark> in shell.tsx replaces the emoji-in-a-gradient-tile badge; the mark takes the honey and the wordmark takes text colour, so the accent lands once. #vault-name sets in Jersey 10 at 18px — the face is condensed, so that measures like 13px of the UI face. - Auth pages (authlocal.go): server-rendered, so they had their own emoji logo. Same mark, inline. - Docs: bear.svg becomes the mark (fixed honey fill — Starlight renders the logo as <img>, which can't inherit currentColor), and .site-title sets in Jersey 10. Starlight tints that title with the accent by default, which put honey on white in light mode and failed contrast; it now takes --sl-color-white, matching the app. - Favicon: the mark, as a data URI. Jersey 10 is SIL OFL and self-hosted in both trees — Vite fingerprints the app's copy into static/assets/, the docs serve theirs from public/ — so no surface makes a third-party font request. Licence ships beside each file. It is deliberately not a design token: tw.css's @theme block is mirrored by the cloud landing's tokens.css and a drift check fails the build if they diverge, so the logo face lives in plain CSS. The cloud landing page carries the same mark and face (separate repo). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
|
b10c6f961e |
docs: a Starlight docs site for docs.beardrive.ai (#38)
Adds web/docs — the public product documentation, built with Astro + Starlight and deployed on its own rather than embedded in the binary. Why a standalone site in the OSS repo, rather than a section of the cloud landing page: - Docs change far more often than the binary does. Embedding them would mean a Go rebuild and redeploy to fix a typo, and would ship a Pagefind search index inside every self-hoster's install. - Self-hosting instructions and the CLI reference document the OSS project, so "edit this page" should resolve to something an outside contributor can open a PR against. - Design tokens get easier, not harder: scripts/tokens.mjs generates src/styles/tokens.gen.css from the @theme block in the hub frontend's tw.css, so there is one source of truth and nothing to police. (The cloud landing sits across a module edge and has to keep a *copy*, guarded by its own check-tokens.mjs.) custom.css maps Starlight's --sl-color-* onto those tokens and invents no colors of its own. Content is seeded from README.md, docs/self-hosting.md, and the plugin skill. Guides deliberately cover agent workflows — connecting an agent, the two-file AGENTS.md orientation pattern, artifacts and links, read heat, scoping the folder — rather than re-teaching the CLI, which lives in Reference. starlight-llms-txt emits /llms.txt at build. Convention wants that at the root domain, so beardrive.ai/llms.txt should point here; that redirect belongs to the cloud landing and is the one cross-repo coordination point this split introduces. Claude-Session: https://claude.ai/code/session_018GcqsM6prjdv9rUrhVVEiC Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com> |