The intake/secretary chat driver treated StreamEvent text deltas as the
ONLY text channel: AssistantMessage TextBlocks were always swallowed as
already-streamed (the double-render guard). The CLI's partial-message
emission turned out to be remotely gated — on 2026-07-23 the NAS
containers got zero stream_event lines from the identical binary,
flags, SDK, model, and settings that stream fine elsewhere — so the
guard became a total blackout: replies were generated, the relay
carried only init/status/turn_end, and the CEO saw nothing.
SdkIntakeSession.send now tracks whether any text delta arrived during
the turn; normalize() emits an AssistantMessage's complete text only
when none did. Streaming mode is byte-identical (deltas render live,
completes stay suppressed); gated mode delivers the reply as one block
instead of nothing. Covers Secretary (same machinery). Regression
tests: fallback emission, default suppression, and both modes
end-to-end through the session layer.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The 0.27.0 release proposal executes its STORED drafted changelog, and
the current proposal predates #663 — approving it as-is would ship a
release whose commit range includes both #663 fixes with no changelog
line, which the next readiness sweep would then flag as curation gaps.
Documenting them in [Unreleased] so a re-originated proposal drafts
complete; verified against the gap-check matching rules (every required
commit in v0.26.0..HEAD matches by #PR or exact summary).
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(providers): Gemini CLI provider — ModelProvider.GEMINI
Mirrors the grok blueprint with source-verified divergences (all facts
pinned against google-gemini/gemini-cli @ 9681621c): no refresher
daemon — Google's refresh tokens are reusable, so the RO host mount is
COPIED to a writable container-local ~/.gemini and each container
refreshes in-process independently (the write-back crash risk on RO
never triggers); settings.json renders security.auth.selectedType
'oauth-personal', experimental.enableAgents=false (subagent ban),
autoConfigureMemory=false with a bounded heap; tool scoping rides the
tiered TOML Policy Engine (deny-only rules that yolo mode structurally
cannot beat); gemini -p with --output-format stream-json; usage parsed
from the run's own stdout stats — the adversarial pass caught the
parser reading the json-mode nested shape while the entrypoint runs
stream-json's FLAT shape (every real run would have priced $0 forever,
hidden by fixtures sharing the assumption) — now flat-primary with the
nested shape as cited fallback; rate-limit classified from structured
error.type only (model-echo immune), native exit 41 auth passthrough;
per-model pricing for the three GA models; migrations 084 (enum) + 085
(seed) complete the 082-085 finale chain. V1 excludes interactive
intake/secretary. Stack-merge required two behavior-preserving
complexity refactors in the shared park/usage plumbing (a park-pair
loop; a usage-reader dispatch dict).
* fix(providers): route gemini usage read through the containment barrier
Mirrors the codex/grok fix — _gemini_usage_json now delegates to
_read_usage_json_contained, so CodeQL's path-injection alert on the
gemini read is resolved by the same resolve-and-contain guard.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(providers): Codex CLI provider — OpenAI via ModelProvider.OPENAI
Mirrors the grok blueprint end to end: CodexCliProvider (RO ~/.codex
mount, ANTHROPIC_* blanked), an orchestrator-side codex_auth.py
refresher (JWT-exp staleness, atomic rewrite, lock-serialized single-use
rotation, --check backstop; the CLI's own in-process refresh write
no-ops on the RO mount by design — margins keep the orchestrator ahead
of the CLI's 5-minute window), config.toml rendering with required=true
gateway MCP servers, execpolicy deny rules (forbidden-only), per-role
--sandbox (developer=workspace-write, review/doc roles read-only),
codex exec --json with pinned ROBOCO_CODEX_CLI_MODEL (gpt-5.3-codex),
usage summed from typed turn.completed events priced via the real
4-bucket split, dedicated image + entrypoint, registry/park/finalize/
compose/release wiring. V1 excludes interactive intake/secretary.
Per adversarial review: migration 083 seeds the openai provider row
enabled=True (without it every routing path 404'd — the whole feature
was operationally dead code; grok needed the same seed in 039), the
panel picker gained the OpenAI catalog group it silently lacked, and
exit classification is structural — only stderr and error.message
fields from error events are sniffed (word-boundaried patterns, exact
auth phrases, bare 'login' dropped), so the model echoing on-topic
words can never false-park the provider fleet-wide, proven by a
benign-transcript test. Known open risk flagged, not claimed: whether
codex's workspace-write OS sandbox excludes /app is unverified, and no
hook mechanism exists to port the bash-guard defense-in-depth.
* fix(providers): containment barrier on usage.json reads (code scanning)
CodeQL flagged the codex usage read as path injection — correctly:
os.path.basename does not neutralize '..', and the upstream segment
validator isn't in CodeQL's taint model. The grok/codex reads collapse
into one _read_usage_json_contained helper that resolves the built path
and refuses anything outside the resolved usage root — a hostile id can
never escape regardless of upstream drift. Traversal + containment
regression tests added; a stray noqa in the test file replaced with a
named constant per repo rule.
* fix(providers): use realpath+startswith containment CodeQL recognizes
The is_relative_to() guard was a real barrier but not in CodeQL's
py/path-injection sanitizer model, so the alert persisted. Switch to
the canonical os.path.realpath + startswith(root + os.sep) form, which
CodeQL recognizes as a path-traversal barrier; behavior is identical
(refuse any candidate resolving outside the usage root).
* fix(providers): regexp-allowlist the usage-id segment (CodeQL barrier)
Neither is_relative_to nor realpath+startswith was recognized by
CodeQL's py/path-injection sanitizer model across the str->Path->open
flow. Sanitize the tainted component at the source instead: the id must
fullmatch a strict slug token ([A-Za-z0-9][A-Za-z0-9._-]*, no
separators, no '..'), which CodeQL recognizes as a path-injection
barrier; the realpath+startswith containment stays as defense-in-depth.
* fix(providers): standalone regexp guard so CodeQL recognizes the barrier
The sanitizer was one disjunct of a compound 'or' condition, which
CodeQL's guard analysis does not trace as a barrier. Split the regexp
fullmatch into its own single-condition guard (the redundant '..' check
is dropped — the required alphanumeric first char already excludes it).
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The 08-31 lever: model_assignments gains one compound rung —
AGENT_SLUG > ROLE('{role}:{complexity}') > ROLE > GLOBAL — so a
low-complexity task can route to a cheaper tier while coordinators stay
pinned. Structurally opt-in: zero rows means byte-identical routing
(pinned by a named test across every precedence case), the cost_tiered
apply-mode (seeds developer:low→haiku) is reachable only from the
explicit PM-gated endpoint — verified no startup path can apply it.
Overrides are downgrade-only (input-price comparator), allowlisted to
{developer, qa, documenter} — cell_pm excluded per the org's own
coordinator definition and its documented weak-model incidents — and
validated at write time (disabled/unconfigured provider rejected with
remediation; cross-provider-family overrides warn explicitly).
Per adversarial review: the four mode-switch applies now spare compound
rows exactly like agent pins (the 2026-07-17 unscoped-wipe class, new
victim, same fix extended via one shared wipe helper) with panel cache
invalidation + truthful confirm dialogs; preset apply validates the
entire payload BEFORE the wipe (validate-all-first), with a savepoint
crash test proving rollback.
Presets (CEO request): routing_presets table (migration 082) snapshots
the full mix — mode, per-agent overrides, complexity rows — with
save/apply/delete endpoints and a panel preset bar; applying skips
since-removed models with per-entry notes, never silently.
Task complexity threads task_id through _resolve_agent_route at both
call sites; taskless spawns unchanged. 235 backend + 23 panel tests.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(notifications): exponential backoff + CAS claim for expired-unacked re-escalation
The sweep re-escalated every expired unacked ack-required notification
on every ~60s tick, forever — the live incident: 3 fresh blocker
escalations + Telegram DMs per minute from a static stale pile. Now
each notification carries reescalation_count / last_reescalated_at /
reescalation_delivered_count (migration 079): first fire at expiry,
then doubling intervals from 1h capped at 24h, hard stop after
ROBOCO_NOTIFICATION_MAX_REESCALATIONS (default 5) with one permanent
log carrying attempts-vs-delivered so 'seen and ignored' is
distinguishable from 'route never worked'. The due/wait/capped decision
is a pure function in foundation/policy/communications.py.
Per adversarial review, the attempt slot is claimed by compare-and-set
(UPDATE ... WHERE reescalation_count = :n) BEFORE delivery — the
previous draft leaned on the 60s dedup window, which never engages for
BLOCKER_ESCALATION (_LOOP_PRONE_TYPES excludes it), so concurrent
sweeps would have double-delivered. A lost claim skips delivery
outright. Legacy rows read as count=0 and keep today's first-fire
semantics. 61 tests incl. a two-session CAS race and a real alembic
upgrade/downgrade round trip.
* feat(budgets): per-task and per-project cost budgets (flag-gated)
tasks.budget_usd + projects.monthly_budget_usd (migration 080, chained
on 079; adds ix_agent_spawn_sessions_task_id since both enforcement
seams filter on bare task_id). Behind ROBOCO_TASK_BUDGETS_ENABLED
(default off, feature-flags card) — verifiably inert when off.
Claim-time: a project-month-spend guard applies to WORK-STARTING claims
only (i_will_work_on / i_will_plan) — per adversarial review, review/
doc/gate/inbound-PR claims are exempt so in-flight work can always
finish reviewing and merging at cap. Spend counts closed sessions'
estimated_cost_usd PLUS open sessions priced live from token snapshots
(the original closed-only sum read parallel long sessions as $0).
Sweep-side: the existing budget sweep also prices the active task's
spend vs budget_usd (TaskType defaults when null); on breach the task
is BLOCKED (HUMAN resolver, budget marker) BEFORE the graceful stop so
the unclaim no-ops and the dispatcher never respawns onto it, and the
CEO notification names both recovery steps. unblock on a budget-blocked
task re-checks live spend and refuses while still over — no silent
re-breach loop. Panel: budget inputs in both dialogs (0 rejected — a
zero budget silently blocks everything), spend logic consolidated in
TaskService.task_spend_usd. 42 new tests incl. a real-DB spend-query
suite and a two-tick non-refire sweep test.
* feat(eval): golden-task eval harness + doctrine cohort stamp
roboco/eval: 6 BenchTaskSpec fixtures run through the real lifecycle in
a disposable environment (the e2e_smoke harness's fake GitHub + local
git origin + throwaway DB catalog — real isolation, not convention),
scored deterministically (terminal status, revision_count, cycle time,
tokens/cost via the agent_spawn_sessions task_id join) plus a local-
model judge whose output is nested under a non_deterministic-marked
object so cohort diffs don't read judge noise as regression. CLI:
python -m roboco.eval run --role <slug> --cohort <name>. Source-
checkout-only by declared posture (deptry-scoped ignore + a hard
ImportError guard naming why; tests/ never ships in images or wheels).
agent_spawn_sessions.doctrine_version (migration 081, chained on 080)
is stamped at spawn-session finalize from the composed prompt layers —
with the session's model column it identifies a cohort durably.
Per adversarial review: bench runs patch the vault flags off (they were
writing real markdown into the operator's vault), and the real-spawn
OrchestratorStageSpawner is deliberately cut to NotImplementedError —
spawned containers' MCP wiring resolves to the production orchestrator
under real agent UUIDs, so real spawns wait for a dedicated follow-up;
the injectable scripted spawner is the working path. Full suite 13852
passed / 94% coverage in the source worktree; deptry/mypy/xenon clean.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(notifications): exponential backoff + CAS claim for expired-unacked re-escalation
The sweep re-escalated every expired unacked ack-required notification
on every ~60s tick, forever — the live incident: 3 fresh blocker
escalations + Telegram DMs per minute from a static stale pile. Now
each notification carries reescalation_count / last_reescalated_at /
reescalation_delivered_count (migration 079): first fire at expiry,
then doubling intervals from 1h capped at 24h, hard stop after
ROBOCO_NOTIFICATION_MAX_REESCALATIONS (default 5) with one permanent
log carrying attempts-vs-delivered so 'seen and ignored' is
distinguishable from 'route never worked'. The due/wait/capped decision
is a pure function in foundation/policy/communications.py.
Per adversarial review, the attempt slot is claimed by compare-and-set
(UPDATE ... WHERE reescalation_count = :n) BEFORE delivery — the
previous draft leaned on the 60s dedup window, which never engages for
BLOCKER_ESCALATION (_LOOP_PRONE_TYPES excludes it), so concurrent
sweeps would have double-delivered. A lost claim skips delivery
outright. Legacy rows read as count=0 and keep today's first-fire
semantics. 61 tests incl. a two-session CAS race and a real alembic
upgrade/downgrade round trip.
* feat(budgets): per-task and per-project cost budgets (flag-gated)
tasks.budget_usd + projects.monthly_budget_usd (migration 080, chained
on 079; adds ix_agent_spawn_sessions_task_id since both enforcement
seams filter on bare task_id). Behind ROBOCO_TASK_BUDGETS_ENABLED
(default off, feature-flags card) — verifiably inert when off.
Claim-time: a project-month-spend guard applies to WORK-STARTING claims
only (i_will_work_on / i_will_plan) — per adversarial review, review/
doc/gate/inbound-PR claims are exempt so in-flight work can always
finish reviewing and merging at cap. Spend counts closed sessions'
estimated_cost_usd PLUS open sessions priced live from token snapshots
(the original closed-only sum read parallel long sessions as $0).
Sweep-side: the existing budget sweep also prices the active task's
spend vs budget_usd (TaskType defaults when null); on breach the task
is BLOCKED (HUMAN resolver, budget marker) BEFORE the graceful stop so
the unclaim no-ops and the dispatcher never respawns onto it, and the
CEO notification names both recovery steps. unblock on a budget-blocked
task re-checks live spend and refuses while still over — no silent
re-breach loop. Panel: budget inputs in both dialogs (0 rejected — a
zero budget silently blocks everything), spend logic consolidated in
TaskService.task_spend_usd. 42 new tests incl. a real-DB spend-query
suite and a two-tick non-refire sweep test.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The sweep re-escalated every expired unacked ack-required notification
on every ~60s tick, forever — the live incident: 3 fresh blocker
escalations + Telegram DMs per minute from a static stale pile. Now
each notification carries reescalation_count / last_reescalated_at /
reescalation_delivered_count (migration 079): first fire at expiry,
then doubling intervals from 1h capped at 24h, hard stop after
ROBOCO_NOTIFICATION_MAX_REESCALATIONS (default 5) with one permanent
log carrying attempts-vs-delivered so 'seen and ignored' is
distinguishable from 'route never worked'. The due/wait/capped decision
is a pure function in foundation/policy/communications.py.
Per adversarial review, the attempt slot is claimed by compare-and-set
(UPDATE ... WHERE reescalation_count = :n) BEFORE delivery — the
previous draft leaned on the 60s dedup window, which never engages for
BLOCKER_ESCALATION (_LOOP_PRONE_TYPES excludes it), so concurrent
sweeps would have double-delivered. A lost claim skips delivery
outright. Legacy rows read as count=0 and keep today's first-fire
semantics. 61 tests incl. a two-session CAS race and a real alembic
upgrade/downgrade round trip.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
scripts/bootstrap.sh: idempotent bring-up for the pull-and-run deploy.
Fresh .env: copies .env.example and injects the three required secrets
using the documented one-liners (the panel token via the exact HMAC
formula issue_panel_token uses), with a standing-credential warning —
and a louder one when cloud auth is detected — since compose's :? guard
refuses an empty token unconditionally (verified live). Reused .env:
never touched, but the three required vars are pre-validated with
pointed remedies instead of compose's opaque interpolation error. Then
pull + up -d + a doctor-style readiness sweep grounded in the real
surfaces (root /health, /api/auth/status through nginx, the verbatim
'Alembic upgrade finished' log line, ollama list), each stage failing
loud with the command to run next. Exposed as make quickstart; README
leads with it and keeps the manual steps as 'what quickstart does'.
Also found and fixed along the way: the documented registry quickstart
was already broken — nginx's :?-required ROBOCO_PANEL_AGENT_TOKEN ships
empty in .env.example, so the 4-step path failed at compose config.
release.yml gains a pull-smoke job (fresh runner, own GHCR login,
needs publish-images): literally pulls the registry compose against the
just-published tag, guarding the missing-image regression class that
already happened once.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Rung protection lived only in delete_task_branch; the post-merge PR-
source cleanup (and the stale-branch sweep's shared primitive) could
still delete a branch that IS a ladder rung. _protected_branches_for_
deletion(slug) — field ∪ rung names, null-ladder shim included — now
feeds _delete_remote_branch_best_effort, so every remote deletion path
is covered; delete_task_branch's local rung check is removed as exactly
subsumed (verified byte-identical comparison semantics). Bonus closed
gap: a renamed trunk (default_branch 'trunk', null ladder) is now
delete-protected, which the hardcoded main/master floor never covered.
Per adversarial review, the deletion lookup fails CLOSED: a raised
project lookup skips the delete with a warning (a skipped best-effort
delete just retries next sweep — free safety), while a genuinely-gone
project proceeds with the hardcoded floor (its ladder is meaningless).
The rebase/sync resolver stays fail-open — a refused rebase on a DB
blip would wrongly block work, a different tradeoff, now documented.
Panel tooltip updated to the new truth. 29 tests.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Two coupled hardenings from the chain-peers adversarial rounds plus the
root-cause fix for the live post-deploy incident where the CEO was
blocked from the panel ('IP not allowed: 100.x.x.x').
Hop peel-set: the whole docker bridge pool leaves the XFF hop set — hops
are now loopback plus operator-named single addresses only
(ROBOCO_GUARD_TRUSTED_CHAIN_PEERS, plain IPs; CIDR entries rejected with
a warning because a range readmits sibling containers). Default-empty
closes the CGNAT-forge residual outright; a gateway-fronted Tailscale
Serve deploy sets its real gateway IP, and a rate-limited detection log
names exactly that IP when an unconfigured host-proxied tailnet chain is
seen, so the silent-regression shape is observable. The connecting-peer
gate (may nginx present XFF at all) deliberately keeps the broad bridge
pool — different check, unchanged.
Incident root cause: guard-core's whitelist is an EXCLUSIVE allowlist
(any non-member is refused), so honestly resolving the tailnet client IP
made ip_security reject the CEO. The tailnet CGNAT range joins
_guard_whitelist() deliberately: Tailscale authenticates device
membership before a packet arrives, real-IP stamping still buys correct
attribution, and any future non-tailnet exposure keeps full scrutiny.
Both compose files now pass ROBOCO_GUARD_EMERGENCY_WHITELIST through to
the orchestrator (the operator escape hatch previously did nothing in a
compose deploy).
NAS is running ROBOCO_GUARD_PASSIVE_MODE=true as interim mitigation —
flip back to false when this deploys. 66 tests.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
projects.protected_branches existed end-to-end but nothing consulted it
— the panel had no editor and the git safety checks used hardcoded sets.
Now: GitService._protected_branches_for(slug) (frozenset, stripped,
fail-open to the hardcoded floor with a warning log) is unioned — never
replacing, only tightening — into rebase()'s refusal set, the shared
_delete_remote_branch_best_effort skip set (threaded through every
caller: task cleanup, PR merge/close cleanup), and sync_task_branch,
which now refuses to force-push a protected-named head (the dev-facing
sync_branch verb path the HTTP-only fix would have missed). Matching is
exact and case-sensitive; an empty list degrades to exactly the old
hardcoded behavior, pinned by union-floor regression tests (master/main
stay refused regardless of the project list).
Panel: chips editor for the field in the edit-project dialog (add via
Enter/comma, paste-splitting on comma-separated lists, dedup, clear-to-
empty persists []) with an honest tooltip scoped to what is actually
enforced. Tests cover both the incumbent GitHub-App dialog suite and the
new Protected Branches suite in one harness.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(x): redraft loop on CEO reject — feedback re-enters the draft flow
A rejected X draft's reason used to die with the cancel. reject() with a
non-blank reason now schedules a redraft after its commit
(defer_after_commit; fresh session; never blocks or fails the HTTP
response): XEngine.redraft_from_rejection re-drafts the same source kind
via the local model with the reason and rejected body folded in as
revision guidance, originating ONE fresh held draft — mirroring the
video pipeline's reauthor_from_rejection. Local-model failure or empty
output originates nothing (no degraded copies); markers carry forward
whole so a redrafted reply/spotlight stays fully functional downstream;
bodies ride the same 280 clamp; the open-posts cap holds.
Hardened per adversarial review: reject() is now idempotent on an
already-CANCELLED target at both check sites (mirroring approve's
already_rejected guard — a replayed reject schedules nothing), and the
dedup check+originate runs under a non-blocking identity-keyed Redis
lock (SET NX + compare-and-del, matching the approve/reject mutex
style) so racing rejects can't stack duplicate drafts — lock held or
Redis down skips the redraft, which is always safe. Tests pin the
fresh-session contract by session identity, the replay no-op, the
lock-skip, and clean up their own committed rows.
* fix(tests): runtime UUID import + typed task-id coercion in x cleanup helper
CI's quality gate runs mypy over tests/ (the local pass covered only
roboco/): the _delete_tasks calls handed ORM-typed ids where uuid.UUID
was expected. Coercing at the call sites then exposed that UUID was
imported under TYPE_CHECKING only — a runtime NameError. Import moved
to runtime; both call sites coerce explicitly.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Root cause first: no viewport export existed anywhere, so viewport-fit
was never 'cover' and every env(safe-area-inset-*) resolved to 0 on
notched iPhones — content under the status bar, dock without real
home-indicator clearance. The export lives on the (tg) group layout
(server component), NOT app-wide: the dashboard shell has no safe-area
padding and must not inherit cover.
Also: min-w-0 on four truncating flex children that overflowed their
justify-between rows (chat names/previews, fleet task titles);
object-contain on the approvals video (letterbox instead of distort on
short phones); break-words on the changelog pre / task description /
quoted mention; touch targets bumped to >=36px (sheet close, segmented
controls, ack button, chips, back button, bell, jump-to-latest,
cut-toggle); overflow-x-hidden backstop on the (tg) main scroller; fleet
avatar strip sliced to 3 with a +N badge instead of silent clipping.
Verified: pnpm typecheck clean, lint 0 errors, panel suite 870/870.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
fastapi-guard peels a fixed trusted_proxy_depth=1 from X-Forwarded-For
(the rightmost entry, which nginx itself recorded). That is correct for
every chain except host-proxied tailnet traffic (Tailscale Serve →
nginx), which arrives as [tailnet-client, loopback-or-bridge-gateway] —
depth-1 resolves it to a whitelisted hop IP, leaving WAF/ban/rate-limit
inert for the whole /tg surface (the documented ceiling).
ClientIpResolutionMiddleware (pure ASGI, wraps SecurityMiddleware so it
runs first) stamps guard_core's request.state.client_ip cache — its
supported pre-resolution seam — for EXACTLY that shape: peel known local
hops (loopback + docker bridge pool) from the right, stamp only when at
least one hop was peeled AND the candidate is in the tailnet CGNAT range
(100.64.0.0/10). Every other shape abstains, so direct LAN clients,
agent containers relaying through nginx (even with forged public-IP
prefixes), and all-hops operator traffic resolve byte-for-byte as
before. Documented residual: a same-bridge container forging a CGNAT
prefix only DE-privileges itself (loses its whitelist exemption). XFF is
read first-occurrence to match Starlette's own header semantics, and a
wiring test pins the middleware mount ORDER, not just presence.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The pg_dump sidecar wrote its dumps to the same disk it protects — one
disk failure lost both. Setting ROBOCO_BACKUP_MIRROR_DIR in .env to a
path on a different disk (external/remote mount) arms a mirror step after
every successful dump: tmp+rename copy, mirror pruned to the same
BACKUP_KEEP, unwritable mirror logs-and-skips without blocking the
primary. Unset, the script never attempts a copy — no fake off-disk
copies on the same disk. Docs gain the mirror setup and a quarterly
restore drill (throwaway pgvector container, pg_restore, row-count
sanity check).
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(lifecycle): inherit advanced upstream base on work re-claims
A re-claim reused a branch cut at an earlier claim, so upstream work
merged since (UX/UI landing on the root after the cell branch was cut)
never reached BE/FE branches — divergence and avoidable conflicts.
_finalize_claim now merges the advanced base into the pre-existing branch
via the dependency-lineage merge: already-ancestor is a no-op, a conflict
aborts at the cut point and leaves a transition note steering the agent
to sync_branch, a clean merge logs an audit trail; never fails the claim.
Double-gated: by role (developer/cell_pm/main_pm — QA/documenter/gate
claims review the branch as pushed and never move it) AND by pre-claim
status (pending/needs_revision only — a PM's i_will_plan re-claim of its
own awaiting_pm_review task must not move a branch that already passed
QA + the PR gate). Fresh cuts already branch from the live remote base.
Cell-PM prompt now orders reading the upstream design docs before
planning.
* fix(lifecycle): extract base-inheritance gate predicate for xenon budget
The four-condition inline gate pushed _finalize_claim to cyclomatic
rank C; the quality gate caps blocks at B. The decision moves to a pure
module-level predicate, byte-for-byte the same logic.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The escalation/approval dispatchers spawn a notification's recipient every
cooldown window for as long as it stays pending. These spawns carry no
task_id, so the PM respawn breaker never sees them — a single wedged
alert/escalation whose recipient never resolves it respawns that recipient
forever. Observed live: fe-pm's unacked alerts kept main-pm/fe-pm spawning
every ~2-3 min for 6+ hours.
Two guards, both gating the spawn after the existing cooldown:
- A hard per-(agent, notification) attempt cap (notification_spawn_max_attempts,
default 5): once a notification has respawned its target that many times
without being acknowledged, stop and log once. The count is id-scoped and
survives map pruning (re-stamp), so a fresh escalation is unaffected.
- A live-work check before spawning: skip when the notification has expired,
is stale past notification_spawn_max_age_seconds (default 6h — wedged or
reloaded from before a restart), or its related task is already terminal.
Fail-open — a failed task fetch or unparseable field never suppresses a
real escalation.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* [943d8c4d] Frontend data freshness and approval-queue reliability audit (#631)
* [233a8b0f] WebSocket reconnect message-loss audit and fix (#625)
* [233a8b0f] fix(panel): add REST catch-up to useNotificationStream on WS reconnect
connection.ts has no message buffering/replay, so a notification published
while the CEO bell's socket was down (disconnected/reconnecting) was lost
forever instead of merely delayed. Add a reconnect-triggered GET
/notifications?unread_only=true catch-up folded into the existing
notification_id dedup so a notification delivered both via catch-up and
live WS is never double-counted, and make clearMessages drop the held
catch-up batch too. use-a2a-live.ts and use-rate-limit-websocket.ts were
audited and already have working reconnect-triggered REST fallbacks
(verified via a2a/page.tsx, rate-limit-banner.tsx, usage-overview-panel.tsx
and their existing F083 tests) so no fix was needed there.
* [233a8b0f] docs(panel): add comprehensive WebSocket hooks reference and reconnect architecture guide
Add panel/docs/frontend/hooks.md with full API reference for useWebSocket, useNotificationStream (with new REST catch-up behavior), useAgentStream, useA2ALiveStream, and useConnectionStatus. Include examples, best practices, and testing guidance.
Add panel/docs/architecture/websocket-reconnect.md documenting the message-loss mitigation pattern: Strategy 1 (REST catch-up for events, used by useNotificationStream) and Strategy 2 (REST invalidation for state, used by A2A/rate-limit consumers), plus the dedup logic ensuring no notification is double-counted on reconnect.
---------
Co-authored-by: Frontend Developer 1 <fe-dev-1@roboco.tech>
Co-authored-by: Frontend Documenter <fe-doc@roboco.tech>
* [d5315683] fix(frontend): add distinct toast feedback for silently-swallowed x-post and release-proposal statuses, plus regression tests for all 4 approval queues (#626)
Co-authored-by: Frontend Developer 2 <fe-dev-2@roboco.tech>
* [cd953838] Data-hook null-guard audit and API client 429 retry-by-method fix (#630)
* [cd953838] fix(panel): gate 429 retry by HTTP method, add hook null-guard regression tests
* [cd953838] chore(conventions): waive test-fixture wrapper in hooks null-guard test
* [cd953838] docs(frontend): document API rate-limit retry behavior and null-guard audit results
Added `docs/frontend/api-rate-limiting.md` to document the 429 retry strategy: GET/PUT auto-retry, POST/PATCH/DELETE require X-Idempotency-Key header. Updated `docs/frontend/hooks.md` to confirm the data-hook null-guard audit found all hooks already have correct `enabled` guards and include a regression test suite for the board-review poll on/off behavior and enabled-guard assertions.
---------
Co-authored-by: Frontend Developer 1 <fe-dev-1@roboco.tech>
Co-authored-by: Frontend Documenter <fe-doc@roboco.tech>
---------
Co-authored-by: Frontend Developer 1 <fe-dev-1@roboco.tech>
Co-authored-by: Frontend Documenter <fe-doc@roboco.tech>
Co-authored-by: Frontend Developer 2 <fe-dev-2@roboco.tech>
* [4534c71a] Backend concurrency, state-machine, and engine audit (#634)
* [41de844a] fix(lifecycle): sync CLAIM_RULES with runtime + clear stale claimant on PM hand-off (#627)
Two confirmed state-machine gaps found while auditing lifecycle.py,
task_lifecycle.py, the _ESCALATABLE_TO_BLOCKED bypass, and every
_REVIEW_QUEUE_STATES entry point:
- lifecycle.py's CLAIM_RULES/claim-ActionSpec/StatusTransition table
did not grant CELL_PM/MAIN_PM re-claim of AWAITING_PM_REVIEW even
though task.py's runtime _ROLE_CLAIM_STATUSES already granted it
and claimed the spec agreed -- the two tables had silently drifted,
breaking i_will_plan re-claim on an awaiting_pm_review task.
- docs_complete's _maybe_advance_to_pm_review pre-assigns a specific
owning PM via assigned_to but left claimed_by/active_claimant_id
pointing at the outgoing documenter, unlike every sibling transition
into a review-queue state. A stale active_claimant_id makes
content_actions.py's _active_claim_violation wrongly reject the
newly-assigned PM's own content writes before it formally claims.
Reassign claimed_by + active_claimant_id to the owning PM alongside
assigned_to.
Adds a regression test asserting the documenter's stale claim does not
survive the docs_complete -> awaiting_pm_review hand-off.
Co-authored-by: Backend Developer 2 <be-dev-2@roboco.tech>
* [0c46f666] Engine dedup race + sequencing.py edge-case audit (#628)
* [0c46f666] fix(sequencing): dedup race audit + collision-edge fallback bug
Audited the list-open-then-originate dedup pattern across six engines:
RoadmapEngine, XEngine.run_cycle, DepUpdateEngine, and CIWatchEngine each
run inside exactly one sequential orchestrator-loop asyncio task (no other
call site invokes run_cycle), so they cannot race with themselves; their
in-cycle dedup sets/keys are correctly built before any commit. SelfHealEngine
is the same shape. VideoEngine.open_video_task is genuinely different: it is
reachable from the release-publish hook, the feature-spotlight hook, and the
on-demand POST /video/request route, so two overlapping calls for the same
occasion can both pass the "no open task yet" check before either commits.
Fixed by wrapping the check+insert in a short-lived Redis mutex (reusing
HeartbeatMutex) keyed by occasion, mirroring XPostService's existing
lock pattern, with a regression test proving only one of two concurrent
calls creates a task.
Verified ReleaseExecutor's half-landed retry path (release_commit_sha):
apply_version_bumps and write_changelog_entry both run as uncommitted
working-tree edits before commit_and_push's single `git add -A` + commit,
so a bumped-version-without-changelog state can never reach origin (and
therefore can never be observed by a fresh retry clone) - confirmed correct
with a real-git-repo regression test, no fix needed.
Fixed sequencing.py's dev_task_collision_edges: the `if edges: return edges`
short-circuit dropped the same-assignee-lane fallback entirely whenever ANY
surfaced sibling pair produced a collision edge, even for a completely
unrelated same-assignee pair with no declared surface. Now the fallback
always runs, skipping only pairs the analyzer already ordered (so the two
mechanisms can never disagree on direction for the same pair).
Verified sequencing.py rule 3 (all-shared batch generates no edges): correct
by inspection (_shared_last_edges skips every pair when both are shared) and
confirmed with a regression test - no fix needed.
* [0c46f666] docs(reference): concurrency audit summary - engine races, fixes, verified patterns
---------
Co-authored-by: Backend Developer 2 <be-dev-2@roboco.tech>
Co-authored-by: Backend Documenter <be-doc@roboco.tech>
* [8f7f167a] Redis mutex pre-lock write audit (#629)
* [8f7f167a] Redis mutex pre-lock write audit: add cross-session regression test for XPostService.approve
Audited x_post_service.py, video_post_service.py, release_proposal.py, and
heartbeat_mutex.py for the pre-lock DB-write anti-pattern (a session write
that happens before the SET NX / HeartbeatMutex acquire returns a token,
letting a losing racer's stale write clobber a winner's committed state).
XPostService.approve, VideoPostService.approve, and
ReleaseProposalService.approve/reject already implement the correct
validate-pure-pre-lock, apply-under-lock pattern (the XPostService fix
already shipped per CHANGELOG.md: "X edited_body write deferred into the
single-flight lock (M5)"). HeartbeatMutex holds no AsyncSession at all, so
the anti-pattern is structurally inapplicable there.
Adds a genuine cross-session concurrency regression test to
test_x_post_service.py (a real second DB connection, not an in-process
mock) mirroring VideoPostService's existing cross-session test, proving a
concurrently-committed post survives and the CEO's edited body never lands
on the just-posted row.
* [8f7f167a] Remove redundant inline comments flagged by QA in cross-session regression test
Both comments restated what the surrounding docstrings already say
explicitly, per QA findings F-dbadd8f0 (line 294) and F-27ac051e (line
631) — no behavior change, tests re-verified green against a sandbox
Postgres.
* [8f7f167a] Remove inline trailing comments flagged by QA (correct file this time)
QA findings F-e6f3e6a6 and F-24189858 cited tests/unit/services/
test_x_post_service.py:294 and :631 across 5 revision rounds, but that
file never contained the flagged comment text — a repo-wide grep for
the exact quoted strings shows both comments actually live in the
mirrored tests/unit/services/test_video_post_service.py file, in its
own cross-session concurrency regression tests (the caption-edit and
tiktok-skip tests). Removed both there:
- "# externally visible to the "concurrent" session below" on the
db_session.commit() call
- "# never attempted without credentials" on the tiktok_poster.calls
assertion
Both restated what the surrounding docstrings/test names already say;
no behavior change. Verified with the full make quality gate against a
sandbox Postgres/Redis: 13,717 passed, 94.41% coverage, clean except
one pre-existing unrelated failure in tests/unit/api/test_cloud_auth.py
::test_login_route_parses_oauth2_form_not_query_params, which connects
to the app's default localhost:5432 Postgres (not the db_session
sandbox fixture) and is unreachable in this sandboxed environment —
structurally unrelated to the auth subsystem this task never touches.
* [8f7f167a] Redis mutex pre-lock write audit (round 7): add cross-session regression tests for reject() lock protection
Round-7 QA findings F-7eb9fbcb, F-06f39a2e, and F-4d56e49b claim
XPostService.reject(), ReleaseProposalService.reject(), and
release_executor._await_proc() lack lock protection / a CancelledError
handler — but their cited line ranges (255-267, 429-454, 241-257)
describe a pre-fix, shorter version of these functions that predates
commit fb293a787d, already on this branch. At current HEAD:
- x_post_service.py reject() (lines 275-299) acquires _LOCK_PREFIX,
re-reads under the lock, applies markers.set_x_reject_reason() +
CANCELLED only inside the critical section, releases in finally.
- release_proposal.py reject() (lines 460-486) does the identical
dance with _RELEASE_LOCK_PREFIX.
- release_executor.py _await_proc() (lines 257-265) already has an
`except asyncio.CancelledError` block that kills + reaps the child
and re-raises, mirroring the TimeoutError handler, with an existing
dedicated regression test
(test_await_proc_kills_child_on_outer_cancellation).
The one genuine gap: neither reject() path had a cross-session
(real second DB connection, not an in-process mock) regression test
proving the in-lock re-read catches a concurrent approve/publish that
completes mid-lock-wait — only approve() had one. Added
test_reject_concurrent_approve_completes_during_lock_wait to both
test_x_post_service.py and test_release_proposal_status_guards.py,
mirroring the existing approve() cross-session test: a second engine
commits COMPLETED between reject's pre-lock read and lock acquisition,
and the test asserts the CANCELLED write / reject-reason marker never
lands on the just-completed row.
No production code changed — verified via 103 targeted tests green
against a sandbox Postgres/Redis, plus `make -o sync gate` clean.
* [8f7f167a] Regenerate stale lifecycle artifacts (restore auditor waive_finding)
foundation-check was the only failing gate: the committed lifecycle artifacts
were missing the auditor's waive_finding verb that the lifecycle source
defines, so make quality regenerated them and failed on the diff — nothing to
do with the mutex fix (which passes ruff/mypy/tests/coverage/bandit clean).
make lifecycle restores the drift; this is what the 8 revision rounds kept
missing.
---------
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* [d615f2e3] fix(tests): sync stale CLAIM_RULES pinning assertions with lifecycle.py (#635)
test_claim_rules_match_pre_gateway_table still asserted the pre-audit
two-member frozenset for CELL_PM/MAIN_PM claim rules. CLAIM_RULES in
lifecycle.py already grants both roles claim rights on
Status.AWAITING_PM_REVIEW (added by the state-machine exhaustiveness
audit) so a PM can re-claim its own review-queue task after a respawn.
Updated both assertions to include AWAITING_PM_REVIEW, matching the
actual dict. Grepped the repo for sibling stale copies of the old
literal; found none beyond this test.
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
* [3c4e7a35] fix(quality-gate): reflow CONCURRENCY_AUDIT.md and stub occasion lock in video tests (#636)
Root cause: PR #634's CI failed at the markdown-reflow check (make quality
Makefile:285) on CONCURRENCY_AUDIT.md — a hard-wrapped audit doc left over
from the merged "Engine dedup race + sequencing.py edge-case audit" unit
(PR #628). Fixed with `make reflow-docs` (the exact remedy the CI output
itself named).
Running the full local `make quality` (with a sandbox Postgres/Redis to
get past DB-gated skips) surfaced a second real regression from that same
PR #628 unit: it added a Redis-backed HeartbeatMutex occasion lock to
VideoEngine.open_video_task, but two pre-existing test files
(tests/unit/runtime/test_video_render_loop.py and
tests/integration/test_video_routes.py) call open_video_task without
stubbing that lock, so they failed closed against the suite's
deliberately-unreachable test Redis (_no_live_redis). Fixed by applying
the same lock-stub pattern tests/unit/services/test_video_engine.py
already uses for its own occasion-lock tests: an autouse HeartbeatMutex
stand-in fixture in test_video_render_loop.py, and wrapping the two
route-level video-request tests in test_video_routes.py with the file's
existing _LOCKED patch pair (already used by every other lock-dependent
test in that file).
The one remaining local failure,
test_cloud_auth.py::test_login_route_parses_oauth2_form_not_query_params,
is a pre-existing environment gap unrelated to this branch: it needs a
real Postgres reachable at localhost:5432 (which .github/workflows/ci.yml
provides as a service container) but this dev sandbox has no such binding
— confirmed unrelated to any of the four merged audit units.
make quality now passes clean: 13729 passed, 0 regressions, 94.49% coverage.
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
* [ddc8121f] regenerate lifecycle artifacts for awaiting_pm_review claim rules and waive_finding intent (#637)
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
---------
Co-authored-by: Backend Developer 2 <be-dev-2@roboco.tech>
Co-authored-by: Backend Documenter <be-doc@roboco.tech>
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(sequencing): drop lane-fallback edges that would cycle against analyzer edges
The dev-task collision fallback unioned the analyzer's authoritative edges
with same-assignee lane edges, deduping only the direct pair. A lane chain
through an unsurfaced middle sibling could still contradict an analyzer edge
transitively (the shared-last migration order inverts plain priority order),
closing a 3-cycle that made add_dependency raise ConflictError and wedged
every later delegate to that parent. Fallback edges are now accepted only
when they can't close a cycle against the edges already kept; a regression
test reproduces the exact scenario.
Also strip pre-merge cruft: remove the root CONCURRENCY_AUDIT.md working
report, delete the near-duplicate websocket-reconnect.md doc, fix the stale
a2a/page.tsx doc citation, correct the api-rate-limiting doc to state
idempotency-key retry is unimplemented, and fix two lifecycle.py comments
that referenced a guard function which never existed.
---------
Co-authored-by: Frontend Developer 1 <fe-dev-1@roboco.tech>
Co-authored-by: Frontend Documenter <fe-doc@roboco.tech>
Co-authored-by: Frontend Developer 2 <fe-dev-2@roboco.tech>
Co-authored-by: Backend Developer 2 <be-dev-2@roboco.tech>
Co-authored-by: Backend Documenter <be-doc@roboco.tech>
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(panel): A2A transcript/list poll as fallback when the socket drops
The desktop A2A view refreshed ONLY on /ws/system a2a.message frames —
refetchInterval was off (only the /tg mini app polled). So when the socket
flaps (the NAS stack flaps often), the open transcript froze: agent replies
never landed and the thread stuck on the last frame received, even though the
messages persisted server-side. useA2AMessages and useA2AConversations now
take a 10s REST poll gated on the live-stream connection — polls only while
disconnected, never when the WS is healthy, so it's a true fallback with no
wasted requests. Backend was fine (get_messages_admin returns the full
transcript incl. CEO interjects; the frame carries the right conversation_id).
* fix(panel): make the A2A poll an unconditional backstop, not disconnect-gated
Live-checked the NAS: /ws/system connections stay open (10 opens / 0 closes in
30m) and events publish — the socket is NOT flapping, so a disconnect-gated
poll wouldn't fire. The real freeze mode is a silent-dead / half-open WS that
still reports readyState OPEN (no client keepalive ping detects it), where
isConnected stays true. So poll unconditionally: 20s while the socket claims
up, 8s once known-down. Guarantees liveness regardless of why a frame didn't
land.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
#621 only let a NEW project bind to a GitHub App installation (the create
dialog's Select repo picker). An already-imported project on a PAT had no way
to re-route to the App. The Edit Project dialog now carries a GitHub App
section: when App creds are configured, it shows the current binding (App
installation vs PAT), reuses the same SelectRepoPicker to bind, and an Unbind
button to revert to PAT (sends explicit null). Hidden/disabled for non-GitHub
providers. Once bound, git ops (commits, PR reviews) are attributed to the
App bot instead of the operator's account.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
A project that checks in generated artifacts (RoboCo's lifecycle renders,
verb tables) drifts whenever their source changes. The agent pre-submit gate
(make gate) omits foundation-check, so drift is invisible at the desk and only
fails on CI's drift gate — a failure with no link back to the task, which made
one live task thrash 8 revision rounds.
New per-project codegen_command (migration 078): run in the task's worktree
right before push, and any drift committed into the same push, so CI never
sees stale artifacts. Fail-open — a broken/timeout codegen command logs and
lets the push proceed (CI's drift gate is the safety net); a null command
(every project without checked-in codegen) is a pure no-op. Hooked at both
push_branch (open_pr's first push, the PR head CI grades) and push_task_branch
(later re-pushes). RoboCo sets codegen_command='make codegen' (a new Makefile
target — the write counterpart to foundation-check's read) via the panel.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The Auditor and PR reviewers get the real DM button (this branch). Secretary
and Intake aren't A2A-DMable — they run their conversation over a live-session
bridge — so they now carry the same chat icon but it navigates to their own
screen instead: Intake -> /prompter, Secretary -> /business?tab=secretary.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(a2a): CEO can DM the Auditor and PR reviewers
A mid-flight PR reviewer or Auditor that's stuck was unreachable — the CEO
had no way to DM them. Both roles now carry dm/read_a2a, so the CEO can open
a 1:1 and they can reply in-thread through the existing CEO-reply path.
Scoped deliberately: the Auditor stays a silent observer to its peers — it
gains no peer-initiation surface (can_a2a_direct routes it through
_check_auditor_a2a, which refuses every initiation target; it can only reply
inside a CEO-opened DM). PR reviewers keep their owning-PM scope. Intake and
Secretary stay excluded — they have their own dedicated chat pages.
NO_COMMS_ROLES drops to {prompter, secretary}; the panel's EXCLUDE_NON_DM_ROLES
matches. KB/docs updated so the 'auditor/pr_reviewer have no dm' claim isn't
left stale.
* test(a2a): smoke guard checks _NO_COMMS_ROLES, not a hardcoded 'auditor'
The dm() runtime guard no longer names the auditor (it now carries dm to
reply to the CEO); it refuses the canonical _NO_COMMS_ROLES set. Assert on
that set so the smoke test tracks the guard, not a stale role name.
* chore(foundation): regenerate verb tables for auditor/pr_reviewer dm+read_a2a
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
formatBucket guessed hourly vs daily from the timestamp string
(bucket.endsWith('T00:00:00.000Z')) and then rendered LOCAL getHours(), so
every daily midnight-UTC bucket rendered as the viewer's local hour — '02:00'
at UTC+2 — across the 7d/30d/90d windows. Granularity is now derived from the
data's own bucket spacing (bucketGranularity: min gap >2h = daily) and daily
buckets render as a short UTC date. The 24h/hourly view is unchanged. Added
minTickGap so the 90d axis doesn't crowd.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(github-app): App credentials, installation tokens, and a Select repo picker
RoboCo was 100% PAT-based. A singleton Fernet-encrypted github_app_credentials
row (migration 077, telegram-credentials pattern) now stores the App id +
private key; github_app_auth mints RS256 app JWTs and caches installation
tokens until 5 minutes before expiry. Projects can bind an installation
(projects.github_installation_id): get_decrypted_token returns a minted
installation token for bound projects and falls back to the stored PAT on
any minting failure, so all ten token consumers work unchanged.
CEO-gated routes expose credentials CRUD plus installation/repo listing, and
the New Project dialog gains a Select repo picker (disabled with a HelpTip
until the App is configured) that fills the git URL and binds the
installation; manual URL + PAT stays the default path.
* test(panel): mock the GitHub App credentials card in the settings page test
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
#615 merged on a false green: the CI paths filter excludes motion/**, so a
motion-only commit fired no quality run and the merge landed two latent
breakages on slave.
- _get_ceo_agent used scalar_one_or_none on role==CEO, which raises
MultipleResultsFound once a second CEO-role row exists. It now pins to the
earliest-created CEO, mirroring the sibling _get_auditor_agent. This is what
made test_brand_voice_nudge_fires_once fail under the full-suite ordering.
- _process_mentions tipped to xenon rank C when the skip-drafting branch was
added; extracted the cap check and the skip filter into two small helpers.
- ci.yml push paths now include motion/** so a motion-only commit can't
false-green the quality gate again.
Regression test: two CEO rows no longer break the lookup.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(x): real voice guide, slop ban, and caption craft for drafted content
Release posts and mention replies were drafted from a one-sentence voice
stub while the reasoning-backed Head-of-Marketing voice guide only reached
the off-by-default spotlight path. The drafting prompts now carry the full
voice rules, a banned AI-slop list, three style exemplars, hook-first
structure, and an under-240-char budget so the 280 clamp never truncates
mid-sentence.
An empty brand_voice now nudges the CEO exactly once (durable
system_settings marker) instead of silently shipping baseline voice forever.
A failed reply draft skips origination instead of shipping 'Thanks for the
mention!'. Video dev prompts and motion/README gain per-platform caption
templates (X: hook + specifics + outro; TikTok: hook + short lines + few
niche hashtags).
* docs(motion): reflow the new Captions section to satisfy the prose gate
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(notifications): task titles and agent slugs replace raw UUIDs
Notification producers interpolated raw task/agent UUIDs into subjects and
bodies ('Task 68e1e4db-... unblocked', 'handed back to 00000000-...-0004').
A tiny notification_text helper (task_display: title-first with a #id8
fallback; agent_display: identity-map slug first, DB lookup fallback) now
feeds every producer: all 13 NotificationService methods, the 7
delivery-service bodies whose subjects were already title-based, the
substitute-PM ad-hoc insert, and the orchestrator/choreographer callers,
which thread the task row's title one call deeper. Fixes the literal
'cell_pm' role string sent as an agent slug in the merge-conflict
notification. Tool-call examples like unblock('<uuid>') keep the raw id on
purpose — agents need it.
* test(notifications): board-review subject assertion matches the humanized format
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(git): branch list classifies remote refs correctly and prunes stale ones
The branches route detected remote-tracking refs via a 'remotes/' prefix
that --format=%(refname:short) never emits, so every origin/* ref rendered
under LOCAL and origin/HEAD surfaced as a fake branch. Listing now uses the
full %(refname) and classifies on refs/heads/ vs refs/remotes/.
Cleanup's remote deletion worked, but no code path ever pruned the viewing
clone's remote-tracking refs, so deleted branches persisted in the UI
forever. The branches route now runs a best-effort 'git remote prune origin'
before listing remote refs, and the manual Fetch fetches with --prune.
* refactor(git): extract branch-line classifier to satisfy the complexity gate
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The header chip and the Settings User Info card rendered a literal 'Renzo'.
The name now lives in the system_settings store under ceo_name (validated:
trimmed, non-empty, max 60 chars) with the same client-served default the
transcript-retention card uses, editable inline from the User Info card.
Agent prompts already refer to 'the CEO' generically, so no prompt rewiring;
the two agent-facing RAG docs drop the name too. License/CLA copyright is
untouched.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
New Project claimed GitLab/Gitea were 'planned' while both providers are
fully shipped, showed a hardcoded GitHub badge, and never sent git_provider
at all — a non-GitHub project could not be created without an immediate
edit. It now carries the same forge Select the edit dialog has; the edit
dialog's own 'GitLab support is planned' tooltip is corrected too.
Also: the git actions panel's hardcoded 'main' (wrong PR-eligibility and
target label for master-default and env-ladder projects) is replaced by the
project's resolved head branch; acceptance criteria become editable in the
edit-task dialog; three feature flags get their missing descriptions; and
three forms swap raw-UUID text inputs for the existing Task/Agent selectors.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The per-agent override grid rendered a hand-maintained AGENT_GROUPS literal
that had drifted: ux-dev-2 and all four PR reviewers were absent, so their
model overrides could not be viewed or edited at all (the backend accepts
any slug). The grid now derives its sections from useAgentDefinitions()
with the same team helpers the Fleet page uses, org-ranked ordering, a
loading skeleton, and group HelpTips. The static offline-fallback maps
(agent-utils, use-agents, mock-data) get the missing agents too.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Every chart hand-rolled its own k-only formatter (22596k-style ticks, left-
clipped y labels), used recharts' default white tooltip (invisible header on
the dark theme), and two charts pinned a numeric XAxis interval that collapses
short series to a single tick. The agent Token Activity chart had no axes at
all and blanked its tooltip date on purpose.
One shared formatTokens/formatBucket (lib/format.ts) and one shared themed
tooltip style (components/charts/chart-tooltip.tsx) now feed all 10 charts;
axis widths/margins sized to the labels; preserveStartEnd tick intervals.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(tg): Mini App V6 — premium overhaul (design system, Chat parity, Metrics drilldown, CEO verbs)
Design system: native type with tabular-numeral heroes (mono demoted to
the wordmark), borderless elevated cards, floating dock, Telegram
window-chrome painting via the theme bridge; Inbox moves behind a header
bell with humanized notifications (UUIDs resolve to task names).
Chat: honest Mine/Fleet split — participant-scoped CEO threads with real
unread counts and mark-read, watched fleet threads with reply-as-CEO on
task-linked conversations (watch-only otherwise), markdown transcripts,
live pulse flashes, and a pinned Secretary live chat on the panel's SSE
session runtime.
Metrics: new tab with period-segmented spend hero, by-agent/team/model
breakdowns, delivery + efficiency health, and a per-agent drilldown over
usage time-series (agent_slug) + member scorecard.
Board: tg-native grouped pipeline replacing the MobileTaskBoard wrapper;
task sheet gains the CEO decide verbs (approve / request changes /
unblock).
Security: /api/dashboard router now require_panel_token-gated at router
level (mirrors /api/usage), closing unauthenticated metrics exposure.
* fix(tg): restore Share Tech Mono brand voice, Phosphor icon set, borderless avatars
The mono returns as the numeral/brand voice (.tg-display — heroes, stat
values, wordmark) while labels stay native sentence case. The hand-drawn
duotone glyphs and lucide feature icons are replaced by Phosphor (MIT):
duotone at rest via an IconContext at the shell, filled weight on the
dock's active tab; row glyph maps (board statuses, inbox kinds, approval
kinds, quick actions) all move over. Team avatar tiles drop their borders
— tint-only squircles.
* fix(tg): fleet avatar strip breathes — spaced tiles instead of overlap
* polish(tg): taste-skill audit pass — em-dash purge, one icon family, separator rationing
Applied the design-taste audit against the cockpit: every em-dash in
visible UI copy rewritten (periods/commas/colons), the remaining lucide
chrome (carets, arrows, send, close, spinners) moved to Phosphor so the
tg tree ships one icon family (send is the native paper-plane, carets
bold), the hand-rolled chevron SVG deleted, and metadata lines rationed
to a single middle-dot separator.
* polish(tg): pipeline chip strip scrolls without a visible scrollbar
* fix(tests): metrics observability fixture uses a relative timestamp
The hardcoded _T0 (2026-06-20) aged out of the service's 30-day window
exactly 30 days later, detonating the suite on every branch. Two days
back from now() stays inside every window (30d metrics, 7d scorecards)
permanently.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* [68e1e4db] feat(video): author release-0.26.0 marketing clip in the panel-demo kit register
* [68e1e4db] fix(video): extend pk-frame clip window to full length to kill black-gap render bug
* [68e1e4db] docs(video): document release-0.26.0 composition structure in motion/README
Added a comprehensive "Release-specific example: release-0.26.0" section documenting the v0.26.0 marketing video composition: a 40-second panel-demo kit clip showcasing four key features (orchestrator security hardening, three-forge support, Telegram Mini App V4, active guard mode) with choreographed panel elements, stats overlay, and dual-format captions. Includes preview/test instructions, props.js shape, captions schema with verified character counts, and smoke-test invariants matching the pattern established by release-0.25.0.
---------
Co-authored-by: UX/UI Developer 1 <ux-dev-1@roboco.tech>
Co-authored-by: UX/UI Documenter <ux-doc@roboco.tech>
A source=video authoring task reaches awaiting_ceo_approval with no MP4
yet — rendering only happens after it completes — so the CEO had nothing
to review. Two CEO-gated routes now serve the request_render preview
frames: GET /video/preview-frames/{task_id} lists them per orientation
(parsed from the self-describing .previews/{task8}/{orientation}/ filenames
rather than the render_preview marker, which only holds the last call's
single orientation), and .../{orientation}/{filename} streams a frame's
PNG behind the existing path-confinement guard. The task-detail Overview
gains a Video preview card — a 9:16/1:1 toggle + prev/next/scrubber frame
stepper with composition id, duration, and a dirty badge — shown for a
video task with preview frames or awaiting CEO approval.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
draft_release_post was fed highlights=list(report.change_summary) — raw
per-commit subjects — so the announcement model parroted the top commit
('RoboCo API v0.26.0 is out: docs: curate the full Unreleased body #601').
New pure changelog_highlights() extracts the bold feature leads from the
curated release entry (report.drafted_changelog), stripping PR refs and
trailing periods; approve() prefers those and falls back to change_summary
only when the changelog yields nothing. The video captions were already
good because the authoring dev read the changelog — same source now.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
With the guard active on the NAS, a documenter's journal-entry POST body
tripped a WAF signature and the guard banned its docker-bridge IP
(172.18.0.7) — after which EVERY gateway verb from that agent (dm,
i_am_idle, claim_review) was blocked by ip_security, wedging the agent
into a respawn loop. Confirmed live: roboco:guard:banned_ips:172.18.0.7
in redis with passive=False. The guard's threat-ban targets the external
attack surface arriving via nginx; internal HMAC-authenticated agents
reach the orchestrator DIRECTLY on the docker bridge and must not be
subject to it. build_security_config now sets whitelist to the RFC1918 +
loopback ranges. External traffic keeps its real client IP (XFF, trusted-
proxy depth 1 — un-spoofable into a private range), so the WAF still
fires on genuine attackers; the middleware tests model that with a
public TEST-NET-3 IP.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The release engine's bell notification for a just-originated proposal
inserted through a fresh session while the proposal task sat uncommitted
in the engine's own transaction — the related_task_id FK rejected the
row and the ping was silently lost (caught live in the postgres log; the
DB-free Telegram DM still went out). send_ack_notification now accepts
db_session, forwarded to _create_notification so the insert joins the
caller's transaction, and the release engine passes its session. The
other five callers pass no task_id or reference committed tasks.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The release proposal's drafted changelog carried only the five entries
PRs had remembered to write, and its own gaps list flagged ~53 missing
items — the entire feature story (Mini App V4+V5, the forge program,
Telegram V2/V3, Agents hub, Workstation, video craft, the panel-perf and
hermetic-test-suite work). The Unreleased body now tells the whole
release: Security amended for #599's calibration (the stale
placeholder-trips-the-guard warning was inverted by the fix) + adm-zip
and registry-auth entries; Added carries the thirteen feature programs;
Changed covers perf/agnosticism/test-hermeticity/docs; Fixed absorbs the
remaining baskets. The readiness drafter prefers this curated body, so
the re-originated proposal ships it verbatim.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The app-level agent-roster sync fires /api/agents on every surface; on
/tg it races the initData sign-in, takes the cloud-auth 401, and the
interceptor bounced the Telegram webview to the password /login page it
cannot complete — hijacking the cockpit into the dashboard. The Mini App
owns its auth UX (initData sign-in + its own wall), so the redirect now
exempts /tg via an exported, tested path predicate.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The 2026-07-19 outage root cause: enforce_https keyed off
environment==production, but nginx is the single entry point — the app
only ever sees proxy-HTTP, and the production NAS terminates no TLS at
all — so the moment the guard went active, https_enforcement blocked
the entire request stream. It is now hardcoded off at the guard-config
level (TLS and http->https redirects belong to nginx, not the app).
Same calibration pass for the two prose validators the flip armed: the
secret-exfil key pattern requires a real b64-shaped value so the
documented placeholder lines can't block, and the injection override
pattern requires the second-person 'your' so neutral engineering prose
about the guard subsystem passes.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* perf(panel): kanban virtualization + row memoization + scorecards batch endpoint
The audit's remaining phases: the kanban card lists render through
@tanstack/react-virtual windows (columns are the dnd drop targets, cards
only drag — no sortable conflict) with memoized columns/cards and a
stabilized handleAction; the task table's desktop row and mobile card are
extracted and memoized (the table itself already client-paginates to 100).
Backend: the per-member scorecard N+1 (~20 requests x 3 queries per poll)
collapses into GET /dashboard/metrics/members backed by
get_all_member_scorecards with grouped rollup/overlay SQL shared with the
single-agent path.
* test(metrics): shared-DB-safe scorecard tests — unique seeds, delta assertions
The two new batch-scorecard tests assumed a private DB: fixed ceo/system
slugs collided with other tests' seeds (ix_agents_slug) and a global
exactly-one CEO lookup + exact roster count broke in the one-process
suite. Unique slugs, subset/disjoint assertions, dead count constant
dropped.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
An operator .env with deploy-flavored values skewed local runs three
ways: environment=production flipped the GHSA-4f7g fail-closed auth gate
(header-trust tests 401), an armed cloud_auth 401'd every agent request,
an armed guard rate-limited the suite, and a missing encryption key
failed every crypto path. The autouse fixture pins environment,
encryption_key (per-process Fernet), cloud_auth_enabled, and
guard_enabled to schema defaults; suites exercising those postures arm
them per-test. Full suite locally: 13641 passed, 0 failed — was 174
failures under an armed .env.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Slice updates cover the forge package + provider parity, Telegram cockpit
V4/V5 surfaces, panel perf (virtualized kanban, scorecards batch route),
close_task_pr_best_effort + dependents guard, NO_COMMS_ROLES + CEO A2A
refusal at conversation creation, notification ack-TTL, docs-site config,
pr_labels base-branch signature, taste-skill prompt layers, prompter
history exclusions, live forge e2e suites, and migrations 075/076.
_complete_map.md is regenerated from the slices — the committed concat
had drifted ~8KB behind its own sources.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Seventeen-page sweep of the agent-facing KB against shipped behavior:
required covers_parent_criteria and per-AC criteria_verified reach the
QA/PM/task-tools pages (the QA docs also named non-callable pass_review/
fail_review — the MCP tools are pass/fail); collision_context lands in the
QA/gate/planning evidence docs; the possibilities matrix gets its own
architecture page + config entry; the auditor page gains its missing
waive_finding and playbook-curation verbs; git-pr-types.md is rewritten
off the long-dead is_root_pr model; PR/workspace/git-error pages stop
assuming GitHub (forge-agnostic + env-ladder semantics).
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(tg): Mini App V5 — brand typography, icon depth, motion, detail sheets
Share Tech Mono (the vendored motion-brand face) becomes the cockpit's
display voice via next/font/local scoped to #tg-shell; icon tiles, circle
actions, and avatars get gradient/ring depth; a dependency-free motion
vocabulary lands (spend count-up, tab rise-in, staggered sections,
sparkline draw-in, sheet slide-up with native BackButton dismiss); the
Board tab gains a tap-through task sheet (ACs, open findings, PR link),
Today's fleet opens a full-roster sheet, and Board/Inbox join the
/tg?demo=1 fixtures.
* feat(tg): custom RoboCo icon set + operations ring
The cockpit stops using stock lucide on its hero surfaces: a hand-drawn
duotone icon set (speedometer, seal, bell, kanban, brand-cursor bubble,
rocket, double-check, broom, robot head) covers the tab bar and the Today
ring. The ring itself stops duplicating the tab bar and becomes real
operations: Ship deep-focuses the release proposal in Approvals, Ack all
bulk-acknowledges pending notifications, Sweep runs the stale-branch
cleanup across every git-configured project behind a confirm sheet, and
Fleet opens the roster.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(security): guard goes active; CEO A2A respects no-comms roles; ack notifications expire
ROBOCO_GUARD_PASSIVE_MODE defaults to false in both compose files — the
deferred post-calibration flip; fail_secure stays off and the env override
remains the rollback. can_a2a_direct no longer short-circuits the CEO past
the no-comms set (auditor/pr_reviewer/prompter/secretary), now canonical
in foundation.policy.communications.NO_COMMS_ROLES and shared with the
content-actions gate; the A2A service refuses at conversation creation
instead of silently suppressing the wake. Ack-required notifications get
expires_at stamped from ROBOCO_NOTIFICATION_ACK_TTL_HOURS (default 48,
0 disables), so the re-escalation sweeper's expires_at query matches rows
for the first time.
* refactor(notification): extract _ack_and_expiry — xenon rank back under B
The expires_at stamping pushed _create_notification_with_session to
rank C; the requires_ack + expiry derivation moves into a helper with
the same semantics and comments.
* test(conftest): dispose the global DB engine after every test
Production code reaching get_db_context()/get_engine() lazily creates the
process-global engine bound to the current event loop; with per-test
function-scoped loops, any later test touching the global path inherits a
dead-loop engine and dies with 'Future attached to a different loop' —
the order-dependent class that has been wandering the suite (cloud_auth
login, metrics, tasks-routes, full-lifecycle) whenever collection order
shifts. An autouse fixture now close_db()s after every test, keeping the
global path loop-local; no-op when untouched.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Thread the deployer's product name through the X reply + feature-spotlight
prompts (B6 leftover; release/video paths shipped in #570); make the
docs-site repo/URL config (ROBOCO_DOCS_SITE_*, defaults unchanged) instead
of a roboco-website hardcode (B8); de-assert our repo from the Main PM
prompt (B10); derive PR labels from the real target branch instead of
literal to-master/to-slave; drop the stale headcount from base.md; and
make the bash-guard's Makefile check require an actual quality/gate/lint/
test target before denying raw package-manager commands (no more
false-remediation loop on Go/Rust Makefiles).
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Task cancellation left the task's PR open on the forge forever: cancel()
now best-effort-closes the recorded PR for the task and its cascaded
descendants (close_task_pr_best_effort resolves owner/repo off git_url —
no clone needed; never raises into the cancel). The bulk stale-branch
sweep gains a dependents guard: a branch still recorded by a non-terminal
task, or serving as a live child's resolve_parent_branch base, is excluded
from the candidate window — mirroring the existing env-ladder-rung skip.
Scoped to the sweep, not delete_task_branch, so the BFS cascade-cancel
can't falsely block a parent's branch on its own about-to-cancel child.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The history-digest pipeline (PR #297) injected past-task data but nothing
told the intake agent what to do with it: prompter.md now carries an
explicit don't-re-propose section (cite duplicates by short id, reference
precedent in notes, let history inform depends_on sequencing), and
list_recent_for_project excludes cancelled tasks so dead work can't pose
as precedent.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The deferred half of the Leonxlnx/taste-skill (MIT) adoption: an
industrial-brutalist / minimalist-editorial / premium-agency aesthetic
vocabulary keyed onto the existing design-bar dials for both FE and UX/UI
team prompts, and a ux_ui-only image-direction section (composition,
palette discipline, anti-slop imagery, mockup conventions) with a pointer
from frontend.md. Layer tests pin presence and team scoping.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Four Phase-0 tests still asserted gitlab.com URLs / git_provider='gitlab'
get rejected, which the GitLab forge provider (#581) made false. They fail
deterministically in isolation and only pass CI when one-process test
ordering masks them — the latent red behind today's flaky quality gates.
Flipped to assert acceptance (auto-detect + explicit), mirroring the GHE
escape-hatch test; unknown-host/unknown-provider rejection tests stay.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(tg): premium Today — spend hero, trend, quick actions, fleet avatars
The cockpit home stops being flat cards and becomes a real app surface:
- Spend HERO: the day's cost at 40px with a signed delta-vs-yesterday
chip and a live 7-day amber area sparkline (hand-rolled inline SVG, no
charting lib in the Mini App bundle).
- Quick-action ring: circular Approve (amber + needs-you badge) / Board /
Inbox / Chat, the wallet-style primary-verb row.
- Needs-you as a rich amber gradient banner (top items + draft chips)
instead of a plain section.
- Fleet as live avatar tokens (stable per-name hue, pulse dot) over the
working list.
- "Shipped this week" day-bars (today emphasized) + week total.
Backend: /telegram/today gains spend.series (7-day cost) + delta_pct and
a velocity series (per-day completed tasks) — two cheap grouped-by-day
queries, same DB-only ethos, degrading to zeros on error.
* feat(tg): color-code approval rows by kind
TgRowIcon gains a tone prop; the approvals list tints each tile per kind
(amber Release / sky X post / violet Video / emerald Roadmap) so a mixed
queue reads as color-coded instead of a monochrome column.
* feat(tg): sender/peer avatars on Inbox + Chat
Inbox notifications and Chat conversation rows adopt the fleet-avatar
language: a per-name-hued initials token leads each card, unread inbox
items carry a subtle primary tint, and both cards move to the rounded-2xl
surface — so every tab now shares one visual system. Board keeps the
shared MobileTaskBoard (already grouped/pill-styled, and reused outside
the cockpit).
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
GitLab's create_org_repo (services/forge/gitlab.py) replaces the Phase-3
synthetic 501 with a real implementation: resolves org (a group's full
path, subgroups included) to a numeric namespace id via GET
/groups/{path}, falling back to the token's own namespace on a 404
(personal-namespace projects); POSTs /projects with the
name/path/description/visibility/initialize_with_readme payload
(visibility mapped private->"private"/"internal"); reshapes the 201
onto the GitHub fields callers read (full_name/clone_url/html_url) and
GitLab's duplicate-path 400 "has already been taken" onto GitHub's 422
shape, text preserved. Gitea's create_org_repo was already real but
untested — added transport-level coverage.
GitHubProvisioningService (services/github_provisioning.py) is now
provider-aware: ROBOCO_PROVISIONING_PROVIDER (github default / gitlab /
gitea) and ROBOCO_PROVISIONING_HOST (self-hosted instance, required for
gitlab/gitea or the service stays disabled exactly like a missing
token/org) pick the target forge; the class/factory names stay
GitHub-flavored for backward compatibility (pitch.py and existing
imports untouched). A shared _is_already_exists() helper recognizes
GitHub's "already exists" (422), Gitea's (409/422 "already exists"),
and GitLab's reshaped "has already been taken" (422). The existing-repo
re-fetch now builds a provider-aware RepoRef (GitLab packs org/name
into the owner field; GitHub/Gitea keep the owner,repo pair). Default
behavior (no new env set) is byte-for-byte the Phase-1 GitHub path,
pinned by a regression test.
Gates: ruff format/check clean, mypy roboco/+tests/ clean (1235 files),
xenon A/A/B clean, targeted suite (forge + provisioning + pitch) 79/79
green.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Caught live on the NAS: every cloud-auth login failed with
422 {"loc": ["query", "db"]} regardless of credentials. Root cause:
auth/manager.py used postponed annotations with a TYPE_CHECKING-only
AsyncSession import, so FastAPI could not resolve get_user_db's
Annotated[AsyncSession, Depends(get_db)] at runtime and silently
demoted `db` to a required query parameter.
The module now evaluates annotations eagerly (no future-annotations,
runtime imports) so an unresolvable annotation is loud instead of a
silent contract change. Regression test mounts the REAL login router —
no dependency overrides, which is exactly why the existing suite never
caught this — and asserts wrong credentials yield 400
LOGIN_BAD_CREDENTIALS, never a 422.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* test(forge): live-GitLab contract suite — verified against gitlab.com
Mirror of the Gitea live suite for GitLabProvider (forge Phase 3):
env-gated (ROBOCO_GITLAB_E2E_URL/_TOKEN + ROBOCO_E2E_SMOKE=1),
self-seeding — creates a throwaway private project under the token's
namespace, pushes real commits, and drives the provider end to end: MR
open → duplicate 409→422 reshape → native source/target filter →
GitHub-shape adaptation (iid→number, head/base, merged) → per-file diff
reassembly → note review → commit-status→check_runs reshape → squash
merge with a mergeability-settle retry (GitLab computes merge status
asynchronously) → branch delete → release (shaped html_url) → the
oauth2 Basic-auth git-CLI claim. Project deleted afterwards.
Green against gitlab.com on first full run — no adapter fixes needed
(the settle-retry is the one live-behavior accommodation). Requires a
token with Project: Create (classic `api` scope, or fine-grained with
project create + API read/write).
* fix(tests): split None-guarded assert so its message can't dereference None
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(forge): Phase 2 — Gitea provider, per-call routing, host registry
Gitea support lands behind the Phase-1 seam:
- GiteaProvider (services/forge/gitea.py): Gitea v1 transport addressed
by instance host (api base from the project's git_url). Where Gitea's
wire contract diverges from GitHub's, the provider adapts responses
back into the shapes GitService already classifies (ShapedResponse):
`token` auth scheme, duplicate-PR 409→422 with the "already exists"
text GitService keys on, commit statuses reshaped into check_runs /
workflow_runs envelopes, APPROVE→APPROVED review mapping, Do-keyed
POST merge, merge-method repo keys, label-color '#' prefix,
client-side head/base PR filtering. Deliberate postures per the spec:
zero-workflows fail-open (statuses-free repo → no_ci_configured) and
merge_branch as a shaped 501 (env-sync cascade lands on missing_ref;
the shared local-git fallback is Phase-2.1).
- ForgeRouter (services/forge/router.py): GitService._forge now routes
per call from RepoRef.host — every existing call site unchanged in
shape. RepoRef gains an optional host; _parse_git_url returns the
host-stamped ref and it is threaded through GitService/release
executor instead of being rebuilt from strings (helpers re-signatured
to take RepoRef).
- Host registry (services/forge/registry.py): in-memory host→provider
map, self-healing — ProjectService.get/get_by_slug re-register on
every read; provider_for resolves gitea projects by git_url host.
- Registration validation now accepts git_provider="gitea"; GitLab
remains recognized-but-rejected. Panel: the read-only Forge badge
becomes a real picker (Auto-detect / GitHub-GHE / Gitea / GitLab
disabled).
Plain git (clone/fetch/push) needs no changes — the Basic-auth
extraheader works on Gitea unchanged. Gates: mypy 392 files, xenon A,
full unit suite 6356 green, integration suite 2257 green.
* feat(forge): live-Gitea contract suite + scheme support + slash-safe refs
Hardening from running the provider against a real dockerized Gitea
1.22.6 (the spec's Phase-2 contract suite, now committed as the
env-gated tests/e2e_smoke/test_gitea_live.py — self-seeding: creates its
own repo, pushes real commits, and drives PR open → duplicate reshape →
list/filter → diff → review → labels → commit-status CI reshapes →
squash merge → branch delete → release, plus a live verification of the
x-access-token Basic-auth git-CLI claim).
Two real findings fixed:
- Branch refs weren't URL-encoded — every RoboCo branch carries slashes
(feature/backend/...), and Gitea's router 404s on the extra path
segments. list_ci_runs + delete_branch_ref now quote the ref
(regression-pinned in the unit suite).
- The API base hardcoded https; a LAN instance serving plain http is a
real deployment shape. GiteaProvider gains a scheme (recorded per host
by the registry from the project's git_url).
ShapedResponse moves to forge/shaping.py (shared by the upcoming GitLab
transport, which needs its text override for diff reassembly).
* feat(forge): Phase 3 GitLab provider + Phase 2.1 local-merge fallback
GitLabProvider (services/forge/gitlab.py): GitLab v4 transport addressed
by host+scheme, subgroup-safe (the MR project path packs into
RepoRef.owner, URL-encoded per call). Adapters translate MR semantics
into the GitHub shapes GitService classifies: iid→number,
source/target_branch→head/base with a merged bool, per-file diffs
reassembled into unified-diff text (ShapedResponse text override,
3-page cap), approve-vs-note review routing (GitLab has no
request-changes verb), pipelines/statuses reshaped into
workflow_runs/check_runs, merge-method repo-key mapping, duplicate-MR
409→422. Reviewer mirroring is skipped (needs numeric ids RoboCo
doesn't store); provisioning stays Phase 4. gitlab.com now auto-detects
at registration like github.com; self-hosted GitLab sets the provider
explicitly (panel picker enabled).
Phase 2.1: neither Gitea nor GitLab has GitHub's server-side merges
API — their merge_branch returns a shaped 501 and
GitService.sync_env_branch now runs the shared local-git fallback
(_local_merge_branch: throwaway clone → ancestor check → merge → push;
a conflict aborts with the remote untouched; same status vocabulary as
the merges-API path).
Also aligns the whole tree with the full gate's tests/-scoped mypy
(provider-test responder typing, e2e_smoke's stale owner/repo shapes).
Gates: mypy 1229 files clean, xenon A, unit suite 6393 green, forge
suites 85 green, panel typecheck/lint clean.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The full-suite CI run executes every test tree in one process against one
database, so other suites' committed rows are visible to the Today-brief
queries and the "empty company" / absolute-count assertions failed
(assert 4 == 0). The suite now snapshots a pre-seed baseline and asserts
deltas, seeds at priority 0 so its rows stay inside the brief's item
caps, and uses a unique agent slug (the fixed be-dev-1 could collide
with leaked rows). Verified by co-running with the known-polluting
integration suites in one process.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
CI runs `mypy roboco/ tests/`; the bridge/cockpit suites were only gated
against `mypy roboco/` locally. Real credentials dataclass instead of
SimpleNamespace, AsyncMock casts where mocks sit behind typed fields, a
return annotation on the fake stream, a None-guard on the consumer task,
and a fresh registry lookup where mypy's literal narrowing read an
assert as always-false.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(tg): P0 — dev mock bridge + Telegram-native foundations
Mini App V4 phase 0. The (tg) shell gains the groundwork every later
phase builds on:
- Dev mock bridge: outside Telegram, a development build falls back to a
no-op WebApp object and skips the webapp-auth POST (the regular panel
session cookie authorizes API calls), so the cockpit is workable in a
plain browser. Production keeps the "Open from Telegram" wall.
- Telegram theme adoption: themeParams map onto the shadcn CSS variables
scoped to #tg-shell (desktop dashboard untouched), colorScheme drives
the dark class, themeChanged re-applies live. Non-hex values are
dropped at the trust boundary.
- Viewport/swipe correctness: shell height rides Telegram's own
--tg-viewport-stable-height (100dvh fallback), vertical swipe-to-close
disabled so list scrolling can't dismiss the app.
- Native chrome bindings: TgWebAppProvider context plus useMainButton /
useBackButton declarative hooks and a null-safe haptics helper —
consumers never touch window.Telegram directly.
* feat(tg): P1 — Today home tab + one-round-trip /telegram/today brief
Mini App V4 phase 1: the cockpit now opens on a "Today" brief answering
"does anything need me?" in one glance.
Backend: GET /api/telegram/today (CEO-gated, rate-limited) returns the
whole brief in one round trip via the new TgCockpitService — needs-you
items (awaiting-CEO + blocked tasks capped for a phone screen, held-draft
counts across release/X/video/roadmap queues), a fleet snapshot with
per-agent current-task titles, today's spend from the day rollup
(degrading to zeros on a usage hiccup, mirroring the CEO overview), and
ship state. Deliberately DB-only: no live GitHub calls, no readiness
snapshot (that path clones), no orchestrator singleton — the CI
red/green proxy is the set of open ci_watch fix tasks.
Panel: TgTodayTab is the new default tab (Gauge icon) — needs-you rows
and draft chips deep-link into the tab that acts on them (with a haptic
tap), fleet/spend/ship render as dense cards, 45s refetch until the P3
WebSocket wiring lands.
* fix(tg): dev mock engages when the CDN bridge loads outside Telegram
Live browser smoke caught it: a bare tab still loads telegram-web-app.js,
so window.Telegram.WebApp EXISTS outside Telegram — just with empty
initData. The dev fallback keyed on a null bridge only, so a dev browser
went down the real-auth path and posted empty initData instead of
mounting the mock. The fallback now treats bridge-with-no-initData the
same as no bridge (a real Telegram launch always carries initData);
production behavior is unchanged.
* feat(tg): P2 — native approvals card stack
Mini App V4 phase 2: the Approvals tab stops stacking the four desktop
queue cards and becomes a phone-native flow — one normalized list across
release proposal / X drafts / video drafts / proposed roadmap items, and
a full-context detail per item:
- Release: version/bump/gate badges, changelog draft, gaps, migration
notes, in-flight + failed-execute banners; approve runs the fail-closed
executor, reject requires a substantive change request (10 chars).
- X: editable body with the live 280 counter, replied-to mention quoted;
approve sends the edited body only when actually edited.
- Video: cut-toggled player (blob-fetched through the authed client — a
bare <video src> would 401), per-platform caption edits with 280/2200
counters; approve sends only checked-in edits.
- Roadmap: the PO's full pitch (description, rationale, ACs); approve
materializes into the backlog per item.
The detail's primary action rides Telegram's native MainButton and back
navigation rides the BackButton, with visible fallbacks outside Telegram
(dev mock, old clients). Haptics fire on outcomes. An acted-on item
vanishes from the refetched queue, popping back to the list by
construction. A failed queue source is surfaced ("list may be
incomplete" / "couldn't load") instead of masquerading as an empty
queue — caught live in the browser smoke.
* feat(tg): dev demo mode — /tg?demo=1 renders canned cockpit data
Development-only: with the flag param present, the Today brief and the
four approval queues resolve typed fixtures (dynamically imported, so
production bundles never carry them) instead of hitting the backend —
the cockpit is fully browsable with zero stack running. Mutations still
go to the real API and fail loudly; it's a showroom, not a simulator.
* feat(tg): P3 — cockpit rides /ws/system live
Chat adopts the desktop A2A invalidate-on-frame idiom over the shared
ref-counted /ws/system socket: every a2a.message frame refreshes the
conversation list and the affected thread, missed-frame gaps are healed
by a reconnect refetch, and the 10s thread poll turns off entirely while
the socket is up (it remains the fallback). The Today brief refreshes on
each USAGE_SNAPSHOT push so the spend line tracks the sweeper live, with
the 45s poll as the socket-down fallback. No new sockets, no backend
changes — the WS gate already accepts the cloud-auth session cookie.
* feat(tg): P4 — deterministic bot command tier + self-syncing menu
Mini App V4 phase 4 (deterministic half): three new bot commands beside
/status /queue /task —
- /agents: who's mid-task right now, from the same TgCockpitService
fleet snapshot the Today brief renders (now public `fleet()`).
- /usage: today's spend from the day rollup.
- /blocked: awaiting-you + blocked tasks, deep-linked into the panel,
capped per section, titles HTML-escaped.
BOT_COMMANDS is the single registry driving /help AND a once-per-process
Bot API setMyCommands sync on the first poll cycle (new client method,
best-effort), so the Telegram command menu can never drift from what the
code implements. The interactive tier (/secretary, /newtask riding a
live Intake interview in-thread) is specced but not in this commit.
* feat(tg): direction-C styling pass — Telegram palette, RoboCo voice
The cockpit stops wearing default-shadcn and gets its own visual
language on top of the P0 themeParams bridge (colors stay CSS-variable
driven, so inside Telegram everything still adopts the user's theme):
- Shared primitives (components/tg/ui.tsx): TgSection grouped cards with
tracked micro-label headers, TgRow list rows (44px targets, press
feedback, 1/2-line clamp), TgRowIcon glyph tiles, TgStat tabular-nums
figures. Every tab composes the same three, so density and rhythm are
identical across the surface.
- Shell renders a centered 430px column (sm:border-x) — the phone UI no
longer stretches across a desktop dev browser.
- Tab bar: tighter type, active stroke-weight shift, backdrop blur.
- Today: needs-you count badge, divided task rows with inline blocked
marker, fleet as mono-named rows, spend/ship as stat tiles.
- Approvals rows as icon-tile cards; detail header gains the kind glyph.
- Inbox/Chat rows aligned to the same card language.
* feat(tg): P5 — /secretary and /newtask live-chat bridges
The bot's interactive tier: both commands bridge the CEO's Telegram chat
into the same in-process runtimes the panel drives — the persistent
Secretary container and the scoped Intake interview.
There is no synchronous send→reply seam (replies land on the session's
single-consumer relay queue), so each bridged session runs one long-lived
consumer task (roboco/services/telegram_bridge.py) that drains
PrompterLiveRegistry.stream and pushes one Telegram message per completed
turn. While a session is live, plain chat text IS the conversation;
/end closes it.
/newtask resolves the intake scope (single project auto-picked, multiple
offered as a tap-to-pick keyboard holding the initial text), and the
interview happens in-thread. A draft proposal renders as a card with
Send-to-Board / Discard buttons: confirm routes through the normal
board-review path (PrompterService.confirm_live_draft, route=board) and
PARKS the session — board feedback later streams straight back into the
same thread, closing the redraft loop from the phone. MegaTask batches
still confirm in the panel only.
The consumer's open stream arms the registry's 60s keepalive, so the
bridge runs its own idle TTL (same setting, parked sessions exempt).
State is per-process in-memory by design (the _PENDING_REPLIES posture);
intake/secretary containers are process-wide singletons, so a bridged
session preempts a live panel session of the same kind by construction.
* feat(tg): cockpit skin — RoboCo dark deck with a constant amber accent
The cockpit no longer inherits the dashboard's white default outside
Telegram: #tg-shell carries its own standing skin (deep slate surfaces,
amber primary) so the Mini App looks like RoboCo everywhere. Inside
Telegram the themeParams bridge now overrides SURFACE tokens only —
background/card/text/hint/border repaint to the user's Telegram theme
while --primary/--ring stay RoboCo amber: Telegram's surfaces, RoboCo's
voice. Demo fixtures also rewritten to neutral content (they previously
depicted unbuilt forge work and already-shipped roadmap items as live).
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* [467e263d] fix(ci): restore green Python quality gate on slave (mypy + xenon)
Two independent bugs broke run 29653535468's 'Python quality gate' job,
not the pydantic-settings pin (already correctly 2.14.2 on this branch).
- git.py: _delete_remote_branch_best_effort's success path fell off the
function with no return, failing mypy's missing-return-statement check
and silently returning None instead of the documented True.
- company_goals.py: CompanyGoalsService.upsert's six repetitive
`if key in data` branches pushed the module's average cyclomatic
complexity past xenon's --max-modules A gate; refactored into a loop
over the mutable-field tuple (behaviourally identical).
* [467e263d] docs(changelog): restore green Python quality gate on slave (CI run 29653535468)
Documented two independent code-level bugs fixed in commit 5ed93ca7:
1. git.py: missing return True statement in _delete_remote_branch_best_effort success path (mypy failure)
2. company_goals.py: cyclomatic complexity refactoring for xenon gate pass
Both bugs identified and fixed by dev team; verified via full CI reproduction locally.
---------
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
Co-authored-by: Backend Documenter <be-doc@roboco.tech>
* feat(forge): Phase 0 — git_provider column + registration-time forge validation
Pointing a project at a GitLab/Gitea git_url used to fail silently, several
steps deep, at first PR. New pure policy module (foundation/policy/forge.py)
detects the provider from the git_url host and validates at the
ProjectService create/update chokepoint: github auto-detects and
auto-stamps, explicit git_provider=github is the GitHub Enterprise escape
hatch, gitlab/gitea are recognized-but-not-yet-supported, unknown hosts get
a loud rejection with guidance. An update changing git_url does NOT inherit
a stored auto-stamped provider (restating the override is required), so a
host swap can't smuggle the escape hatch past validation. Migration 075
adds the nullable projects.git_provider column; the panel project dialogs
show the detected forge. Phase 0 of the forge-providers spec.
* feat(forge): Phase 1 — GitProvider seam, GitHub transport extracted
roboco/services/forge/: base.py holds the pure contracts (RepoRef +
GitProvider ABC, stdlib-only — a later GitLabProvider is implemented by
reading this file alone), github.py the httpx transport (20 endpoint
methods behind one shared request-plumbing helper set), registry.py the
wiring (git_provider column -> provider, failing loud on gitlab/gitea).
GitService keeps its exact public surface and all response
classification; its 26 inline REST call sites route through a lazy
_forge property (several suites build GitService via __new__, so an
__init__-set attribute breaks them). github_provisioning and
release_executor ride the same provider. Zero behavior change — the
pre-existing suites pass unmodified; per-project provider resolution
lands with the second provider.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Item B+C of the video/X per-project targeting spec, plus the
company_goals.company_name field they depend on (migration 075).
- CompanyGoalsService.resolve_product_name is the single fallback chain
(project name -> charter company_name -> RoboCo); XEngine and
VideoEngine both call it and their prompt builders are pure functions
taking product_name — release posts/videos stop hardcoding RoboCo.
- The X and video queue responses carry project_slug/project_name via one
shared unloaded-guard helper (api/schemas/project_fields.py); both
panel queues render a shared ProjectBadge so multi-project drafts are
tellable apart.
- Business -> Goals editor gains the company-name input.
- Fixes a pre-existing test-isolation leak: the company-goals routes test
commits the charter singleton into the session-scoped test DB and
polluted later suites; it now deletes the row on teardown.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(forge): Phase 0 — git_provider column + registration-time forge validation
Pointing a project at a GitLab/Gitea git_url used to fail silently, several
steps deep, at first PR. New pure policy module (foundation/policy/forge.py)
detects the provider from the git_url host and validates at the
ProjectService create/update chokepoint: github auto-detects and
auto-stamps, explicit git_provider=github is the GitHub Enterprise escape
hatch, gitlab/gitea are recognized-but-not-yet-supported, unknown hosts get
a loud rejection with guidance. An update changing git_url does NOT inherit
a stored auto-stamped provider (restating the override is required), so a
host swap can't smuggle the escape hatch past validation. Migration 075
adds the nullable projects.git_provider column; the panel project dialogs
show the detected forge. Phase 0 of the forge-providers spec.
* fix(panel): mock-mode forge detection extracts the real host
CodeQL js/incomplete-url-substring-sanitization: the substring check
matched github.com anywhere in the URL. Extract the hostname (URL parse
or scp-form regex, mirroring forge.py) and require an exact match.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* [e530aa5e] Diagnose and fix roboco-api CI failure (run 29629255153) (#561) (#562)
* [e530aa5e] fix(tests): narrow None before indexing validate_init_data() result in telegram_initdata self-check
CI run 29629255153 failed on mypy, not the historical pydantic-settings
issue (uv.lock already pins 2.14.2). The __main__ self-check block in
test_telegram_initdata.py indexed the dict[str, object] | None return
of validate_init_data() without narrowing away None first.
* [e530aa5e] docs(qa): document CI fix for mypy type narrowing in telegram_initdata test
Explains the root cause (mypy type error in __main__ block), the solution (None narrowing before indexing), and the safe pattern for future test self-checks that call functions returning optional types.
---------
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
Co-authored-by: Backend Documenter <be-doc@roboco.tech>
* [0884b737] Diagnose and fix Python quality gate + e2e lifecycle smoke CI failures on PR #563 (#564) (#565)
* [0884b737] fix(tests): isolate ROBOCO_SDK_URL for scripted e2e-smoke agents
tests/e2e_smoke/harness.py already isolates ROBOCO_AGENT_TOKEN from the
host environment (the #503/#504 fix) but left ROBOCO_SDK_URL leaking
through. flow_server/do_server both default it to
http://localhost:9000 and forward every rejection there for the
per-verb circuit breaker; inside a real spawned agent container that
port is a live SDK loopback, so the breaker records genuine attempts
for the ephemeral test-agent IDs and trips circuit_open mid-test
(test_sandbox_on_demand.py::test_request_sandbox_guard_chain_over_real_api,
which deliberately causes 3 rejections in a row). Point it at a
guaranteed-refused loopback address so every environment gets the same
fail-open bypass a bare CI runner already gets by having nothing
listening on 9000 at all.
* [0884b737] docs(changelog): document e2e-smoke harness ROBOCO_SDK_URL isolation fix
Document the fix that isolates ROBOCO_SDK_URL in the ScriptedAgent harness to prevent the per-verb circuit breaker from leaking state into ephemeral test-agent identities when the e2e-smoke suite runs inside a live agent container. This ensures the suite passes consistently regardless of whether it runs on bare CI or inside a spawned agent.
---------
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
Co-authored-by: Backend Documenter <be-doc@roboco.tech>
* [3b9a1771] Diagnose and fix ALL make quality + e2e-smoke stage failures on PR #563; confirm real CI green (round 3) (#566) (#567)
* [3b9a1771] fix(e2e-smoke): match real embedding dimension when seeding fake journal chunk
test_c3_deleted_journal_unindexed inserted a 4-dim placeholder vector
into chunks_journals, but the e2e stack's app lifespan eagerly creates
that table with the real settings.embedding_dimensions (1024) before
the test runs, so the insert failed with "expected 1024 dimensions,
not 4". Derive _SMOKE_DIM from settings.embedding_dimensions instead
of a hardcoded constant so the seeded vector always matches the
table's actual column width.
* [3b9a1771] docs(qa): document e2e-smoke embedding dimension fix in round 3 CI diagnosis
Recorded the root cause, solution, and pattern for the final e2e-smoke test failure found in comprehensive sandbox testing: the test seeded a 4-dim placeholder vector but the app's eager lifespan init created chunks_journals with the real 1024-dim embedding column. Updated _SMOKE_DIM to derive from settings.embedding_dimensions instead of a hardcoded constant.
---------
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
Co-authored-by: Backend Documenter <be-doc@roboco.tech>
---------
Co-authored-by: Backend Developer 1 <be-dev-1@roboco.tech>
Co-authored-by: Backend Documenter <be-doc@roboco.tech>
* feat(panel): Journals joins the Agents hub as its third tab
* docs(map): Journals tab on the Agents hub; registry retargets
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(panel): Agents hub — Fleet and Conversations tabs, /a2a redirect, DM quick-action
* fix(panel): validate deep-linked DM targets against roster and exclusions; re-arm the dm latch
* docs(map): panel entries for this wave
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(panel): customizable Quick Actions on Overview — registry, defaults, persisted picker
* fix(panel): legacy bar actions join the quick-action defaults; empty state; persist-contract test
* docs(map): panel entries for this wave
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(panel): Workstation card grids with persisted view toggle and sorting
* fix(panel): stable multiplier sort with cell-label fallback after adversarial review
* docs(map): panel entries for this wave
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(motion): demo-register visual design bar, vendored craft references, catalog vocabulary index
* fix(motion): correct design-bar composition claims after adversarial fact-check
* docs(map,rag): design-bar, vendored references, and catalog index on the video-engine surfaces
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(panel): Workstation page — Products and Projects as tabs behind one sidebar entry
* docs(map): workstation page, extracted views, and the local-filter scroll trade-off
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(panel): session-start, 7d overview spend, and 30d business spend charts
* feat(panel): surface work sessions as a Git page tab (route was orphaned)
* feat(git): reap spent task branches and render previews at lifecycle chokepoints; guarded stale-branch sweep
* feat(panel): stale-branch cleanup button on the Git page
* fix(git,panel): cursor-resumable sweep, force-delete spent refs, local filter state
* docs(map,rag): branch/preview reaping, cleanup sweep, git-tab work sessions, wave-2 charts
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
A draft targeting a platform with no credentials could never complete:
the X leg posted and committed its id, the TikTok leg failed as
'transient', and the card sat pending in the CEO queue with retry
semantics that structurally cannot succeed. Unconfigured platforms are
now an explicit skip: the draft COMPLETES when every configured
platform has posted (detail names what was skipped), an approve whose
targets are all unconfigured refuses loudly instead of silently
completing, and genuine post failures keep their partial/retry
semantics. Re-approving a parked card clears it without re-posting
(the already-posted guard is covered by a dedicated test).
Verified: 32/32 test_video_post_service (3 new: skip-and-complete,
all-unconfigured refusal, re-approve recovery), ruff/mypy/xenon clean.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Playwright, the taste-skill design bar, and hyperframes were all
implemented — and each stopped one hop short of the hands doing video
work: playwright reached only fe-qa/ux-qa, the design bar's web-UI dials
actively pointed a video task at 'dense product UI -> motion 2-3', and
hyperframes' agent-facing doctrine never reached any agent. Three wires:
- vendor the official HyperFrames agent skills (hyperframes-core,
-keyframes, -creative) under motion/skills/ at a pinned upstream
commit (Apache-2.0, attribution headers; prose reflowed to house
style, re-vendor note in each header); README and the dev video
prompt block point at them
- register the playwright MCP for a ux-dev spawned onto a source=video
task (_is_video_authoring_spawn: fail-closed role/team/task-source
probe) so the composition author can watch their HTML live in a real
browser between renders — gating-only, agent-ux already bakes the
browser; QA gating unchanged, be-qa/ordinary ux-dev still excluded
- design bar video-mode override in the ux_ui team prompt: video tasks
are films, the web dials do not apply — use the cinematography bar
and the vendored doctrine instead
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
The release cuts read as static screen recordings: one 6s cursor glide
that then vanished, a locked-off camera, metronomic flat card fades. And
two renderer defects were silently wrecking every cut:
- @hyperframes/producer floated (^0.7.36, no lockfile — image builds get
whatever is latest): 0.7.60 fails EVERY render with "Cannot access 'rt'
before initialization". Pinned 0.7.36 exact and committed
package-lock.json so sidecar builds are reproducible.
- the producer's per-clip visibility scheduler runs on a clock that lags
~50% behind the encoded timeline on a 40s cut — the final frame showed
the authored ~18s state, so the tail scenes (bell card, toast, outro)
were silently missing from the MP4. This, not authoring, is why
rendered cuts kept losing their late scenes. Fix: clip windows are for
structural layers only (hero, panel frame, status); every beat rides
base-hidden styles + delayed CSS animations, which run on the true
clock. Verified frame-by-frame: all four cards, stats, toast, and
outro now land exactly on schedule in both cuts.
Craft, made reusable in the kit instead of per-composition heroics:
- kit.js choreographCursor: data-waypoints="t x y [click]; ..." generates
a multi-leg eased path with fade in/out, an idle-hand sway, and click
rings + glyph press dips — the cursor behaves like a hand, never pops
in, freezes, or blinks out
- kit.js choreographCamera: data-shots="t x y scale; ..." — push-ins
toward each beat's focal point, pull-backs for reveals, settle to end
- release-0.25.0 both cuts re-choreographed: the cursor is the CEO's
hand (settle on the intake while it types, ONE submit click, witness
each card completing, acknowledge the toast, exit off-frame); the
camera lives on every beat; springy card entrances replace flat fades
- motion/README.md gains 'Cinematography & rhythm' (shot-list-first, no
locked-off camera, verify motion with frame PAIRS) + the clip-window
rule; kit/README.md documents both engines; the dev video prompt block
carries the craft bar
Verified end to end through the real sidecar: both cuts render green,
32-frame strips read visually — cursor travels and clicks on schedule,
camera moves, every scene present. motion pnpm test 15/15.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* feat(gateway): delegation detail-fidelity — details survive hand-off, both directions
Details thinned out at every delegation hop: a PM child task mapped to no
parent criterion was legal (coverage only surfaced at submit_up, after the
whole wave ran — a 12-subtask docs tree grew through 8 review rounds that
way, one child titled 'docs page and route wrapper' shipping only the
page), and QA could pass work on a gestalt read (a 4-scene video brief
shipped 3 scenes past every gate because the features existed only in
prose). Three chokepoint gates:
- delegate (down): every child must declare covers_parent_criteria
resolving against the parent's real acceptance criteria — no mapping or
an unresolvable ref rejects naming every offending child and the valid
criteria; the success envelope carries parent_ac_coverage
{covered, uncovered} so a wave-planning PM sees remaining gaps in the
same turn. Full coverage stays enforced at submit_up (waves stay legal).
- pass_review (up): mandatory criteria_verified — one {criterion,
evidence} entry per task AC, matched by the findings ledger's
id-or-exact-text matcher, evidence soup-checked and capped; rejects
naming the unverified criteria; entries render deterministically into
qa_notes as '[AC] <criterion> — verified: <evidence>' lines. The old
count-only ac_verdicts gate is superseded (arg kept for back-compat).
- video briefs (structured detail at origination): an enumerable feature
list (release highlights, or input_props.highlights carried onto a
reject re-author) becomes its own scene acceptance criterion, bounded to
the AC caps; a re-author without highlights carries the
feedback-addressed criterion instead.
Extracted findings.py's criterion matcher into shared unmatched_criteria /
uncovered_acceptance_criteria instead of duplicating it; criteria_verified
joins the WAF free-text exclusion set like findings/issues.
* fix(gateway): break the block/unblock wedge — four hardening fixes from the live PM loop
A cell task looped fe-pm/main-pm block/unblock for hours (10 cycles, 43
spawns): a transient GitHub API error resolving CI became an unwaivable
blocker finding whose own fix text said no code change was required, the
submit freshness guard then demanded a commit no finding called for,
escalate_up auto-blocked, and main-pm's correct recovery plan 422'd on
the approach length cap, degrading it to a bare unblock. Four fixes:
- pr_pass CI-unresolvable refusal is now explicitly transient-worded:
retry pr_pass shortly, do NOT pr_fail over a CI-status lookup error —
a platform blip is not a code finding
- submit freshness guard grants ONE unchanged-head resubmission per
head sha when the findings ledger has zero open rows (all addressed
without code changes) — stamped via the resubmit_unchanged_head
marker so the same head can never loop a second time
- unblock carries a flip breaker: block_flip_count marker, and at the
third flip a one-shot CEO notification flags the task as structurally
wedged (unblock itself still succeeds — the breaker signals, it does
not wedge recovery)
- i_will_plan's approach cap truncates at 800 chars instead of
rejecting — an over-detailed plan must never cost the PM its turn
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(release): CI wait polls the prod rung; escape the header tooltip apostrophe
get_latest_ci_conclusion defaults to the ladder's head rung, so
wait_for_ci searched slave for a release commit that lives on master
and timed out after 40 minutes with the run already green. The wait
now passes the prod branch explicitly. Also fixes the
react/no-unescaped-entities error that turned master's Panel CI red.
* fix(panel,video): dead dialog triggers behind tooltips; dotted composition ids render
HelpTip nested inside a Dialog/AlertDialog trigger puts the trigger's
click handler on the Tooltip root, which renders no DOM — the agents
Spawn item and the KB Reindex-All / Delete-index confirms were dead.
Tooltips now wrap the triggers. The video renderer accepts interior
single dots in composition ids (release-0.25.0) with '..' still
unrepresentable, and propose_video refuses an unrenderable id at
authoring time.
* fix(dispatch): restart-safe PM review turns
A leaf task in awaiting_pm_review had no periodic pickup: the closure
dispatcher bailed on childless tasks and skipped PR-bearing review
tasks as already-promoted, assuming the submit-time PM session was
still alive — an assumption every restart breaks. Proven live on the
docs-sync leaf after the 0.25.0 redeploy, which also dependency-blocked
its sibling dev task. Childless awaiting_pm_review tasks now flow to
the PM's review turn, and the merge turn respawns its PM when none is
active.
* feat(video): verify the rendered artifact, not the source
The 14s release-0.25.0 cut shipped with only one of four scenes visibly
registering: the dev authored DOM, the smoke asserted DOM, QA read code —
nobody consumed the rendered MP4 before the CEO did. Close that loop, and
the reject loop behind it:
- sidecar frames mode: POST /render with frames=1..32 renders the cut,
ffprobes the REAL duration, extracts midpoint-sampled keyframe PNGs
(timestamps in filenames), streams a tar.gz back with X-Video-Duration
- request_render do-verb (developer/QA, request_sandbox's shape): renders
the caller's ACTUAL composition — dev's own worktree (head_sha/dirty
provenance), QA a read-only git-archive export of the assembled branch —
extracts frames to the container-shared .previews/ path, stamps the
render_preview marker, returns the paths as envelope evidence
- gate: i_am_done on a source=video task refuses without a stamped
render_preview (Requirement.RENDER_VERIFIED; canonical source string
moved to foundation as markers.VIDEO_TASK_SOURCE; mirrored in the
possibilities-matrix fast path so it cannot bypass the check)
- QA claim_review evidence carries video_context (composition id, the
dev's preview, a re-render instruction) so review checks output
- dev spawn prompt block + a 4th authoring AC order Read-every-frame
verification before submitting
- reject -> re-author: a CEO reject with a reason opens a fresh authoring
task carrying the verbatim feedback + a revise-in-place pointer at the
existing composition (best-effort, never fails the reject) — rejection
feedback no longer dies on the cancelled draft
E2E: rendered the committed release-0.25.0 composition through the new
frames mode locally — the returned keyframes show exactly the reported
failure (blank frame at 5.8s, only 'Env ladder' by 12.8s), the check the
fleet was missing.
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
* fix(release): CI wait polls the prod rung; escape the header tooltip apostrophe
get_latest_ci_conclusion defaults to the ladder's head rung, so
wait_for_ci searched slave for a release commit that lives on master
and timed out after 40 minutes with the run already green. The wait
now passes the prod branch explicitly. Also fixes the
react/no-unescaped-entities error that turned master's Panel CI red.
* fix(panel,video): dead dialog triggers behind tooltips; dotted composition ids render
HelpTip nested inside a Dialog/AlertDialog trigger puts the trigger's
click handler on the Tooltip root, which renders no DOM — the agents
Spawn item and the KB Reindex-All / Delete-index confirms were dead.
Tooltips now wrap the triggers. The video renderer accepts interior
single dots in composition ids (release-0.25.0) with '..' still
unrepresentable, and propose_video refuses an unrenderable id at
authoring time.
* fix(dispatch): restart-safe PM review turns
A leaf task in awaiting_pm_review had no periodic pickup: the closure
dispatcher bailed on childless tasks and skipped PR-bearing review
tasks as already-promoted, assuming the submit-time PM session was
still alive — an assumption every restart breaks. Proven live on the
docs-sync leaf after the 0.25.0 redeploy, which also dependency-blocked
its sibling dev task. Childless awaiting_pm_review tasks now flow to
the PM's review turn, and the merge turn respawns its PM when none is
active.
* [1dae04a7] Revise release-0.25.0 composition to 40s scene-based pacing with four feature cards
* [1dae04a7] docs(motion): update release-0.25.0 README section for 40s four-card revision
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Co-authored-by: UX/UI Developer 1 <ux-dev-1@roboco.tech>
Co-authored-by: UX/UI Documenter <ux-doc@roboco.tech>
* fix(release): CI wait polls the prod rung; escape the header tooltip apostrophe
get_latest_ci_conclusion defaults to the ladder's head rung, so
wait_for_ci searched slave for a release commit that lives on master
and timed out after 40 minutes with the run already green. The wait
now passes the prod branch explicitly. Also fixes the
react/no-unescaped-entities error that turned master's Panel CI red.
* [2f806123] feat(motion): add release-0.25.0 composition extending panel-demo register
* [2f806123] fix(scope): revert out-of-scope backend and panel changes from video branch
* [2f806123] fix(scope): revert out-of-scope backend and panel changes from video branch
* [2f806123] fix(scope): restore out-of-scope files from current origin/master after stale-master revert
* [2f806123] fix(scope): restore motion/pnpm-workspace.yaml from origin/master
* [2f806123] docs(motion): add release-0.25.0 composition example to README
---------
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Co-authored-by: UX/UI Developer 1 <ux-dev-1@roboco.tech>
Co-authored-by: UX/UI Documenter <ux-doc@roboco.tech>
Pure derive_pr_labels (foundation/policy/pr_labels.py) maps a PR's shape
to a stable org-structure label set: to master/to slave (is_root_pr
discriminator), root, MegaTask, and the owning layer (main-pm /
cell/{team} / subtask/{team}). Mirrors batch.py: object|None inputs,
enum-or-string normalization, no DB/I/O. Full slave-targeting semantics
(base_branch vs default_branch) land with the slave/master wiring (W-H);
YAGNI now.
GitService._apply_pr_labels posts the result to the GitHub labels API
best-effort (create-before-add, swallow 422/409, never raises) so a label
failure can never block PR creation. Wired at all three PR-opening sites:
create_pr (gateway path), create_pull_request (REST/task path), and
_push_and_open_conventions_pr (static chore label). Existing PR tests
mock _apply_pr_labels so they never hit the real labels API.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
Reject cancelled the proposal's status but never moved it out of the
held-proposal set, so the one-open-proposal dedup blocked the release
manager from ever re-assessing — a rejected proposal deadlocked the cycle.
reject() now sets CANCELLED (mirroring video_post_service), which
list_open_release_proposals already excludes, so a fresh proposal can
originate next cycle.
A failed ~40min background execute (gate red, CI red, or an unexpected
crash) left the proposal silently PENDING with no signal to the CEO.
_run_approve_background now writes a release_execute_outcome marker
(status + detail) on every terminal outcome, and an 'error' marker on an
unhandled exception. GET /proposal surfaces execute_status / execute_detail
/ execute_in_flight (derived from the in-memory _INFLIGHT_APPROVES registry)
so the panel can show a running badge, a failure block with the reason, and
a Retry-approve label instead of a silent wait.
Co-authored-by: Renn F <rennf93@users.noreply.github.com>
