Carry the original thread root separately from the triggering message so generated repository activity remains visible in the conversation that requested it.
Fixes#2896
**Root cause:** `arboard = "3"` compiles with default features only
(`image-data`), so the Linux build has no Wayland backend. In a Wayland
session every copy lands on the XWayland clipboard; the compositor only
mirrors X11 selections while an XWayland window holds focus, and
arboard's clipboard client is windowless, so Wayland-native apps read
nothing. `set_text` still returns `Ok`, which is why the invite dialog
shows "Copied" with an empty clipboard.
**Fix:** enable arboard's `wayland-data-control` feature. Under Wayland
it uses the `zwlr-data-control` protocol; when `WAYLAND_DISPLAY` is
absent it falls back to X11 as before. The `wl-clipboard-rs` dependency
is target-gated inside arboard, so macOS and Windows builds are
unchanged.
**Verified:** `cargo check` passes on the desktop crate; wl-clipboard-rs
now resolves in the lockfile.
Signed-off-by: Shawn Yeager <shawn@shawnyeager.com>
Refs #2062
This carries forward the relay-mesh recovery work from #2304 by @Bartok9
(cherry-picked with original authorship/sign-offs) and adds the
startup/readiness supervision and packaging fixes found while validating
it against a real two-machine Buzz setup.
## From #2304
- Watch the local OpenAI ingress (`:9337`) after launch and re-arm a
stale relay-mesh runtime.
- Require consecutive failed probes before eviction so a transient
inference stall does not cause a cold restart.
- Bound stale-runtime shutdown, preserve runtime identity across the
asynchronous probe, and never evict a concurrent replacement.
- Re-arm only for agents that are actually running; deliberately stopped
agents stay stopped.
- Persist an actionable sentinel error under the managed-agent store
lock and clear only that error after recovery.
- Treat serve-to-client fallback as an intentional fail-safe; configured
serve restoration remains on its existing path.
## Added here
- Use the inference ingress (`:9337`), rather than management port
`:3131`, as Buzz's client-readiness boundary. A usable client no longer
fails or holds agent-save open merely because management startup is
still pending.
- Supervise the embedded SDK startup asynchronously, publish a pending
status while management is unavailable, and keep its mesh identity
alive.
- Avoid racing a replacement while SDK startup still owns the embedded
runtime. If that pending startup later loses ingress while a running
agent still needs it, request a controlled Buzz restart to reclaim the
otherwise-unreachable SDK thread.
- Defer roster-driven replacement while client management startup is
pending.
- Keep post-launch recovery in a dedicated module so the mesh entry
point remains within the desktop file-size gate.
- Explicitly mark generated Unix sidecars executable. On macOS, copying
over an existing non-executable destination preserved its old mode,
causing packaged `buzz-acp`, `buzz-agent`, and tool sidecars to be
reported as missing.
## Validation
Automated:
- `just ci` — passed, including formatting, Clippy with warnings denied,
desktop/web/mobile checks and tests, and builds.
- Full Tauri mesh-feature suite — 1,702 passed, 0 failed, 15 ignored.
- Mesh-feature Clippy with `-D warnings` — passed.
- Release macOS app bundle with `mesh-llm` — built successfully; every
bundled sidecar passed executable-mode and deep code-signature
verification.
Live two-machine E2E:
- M5: released Buzz serving `unsloth/gemma-4-26B-A4B-it-GGUF:UD-Q4_K_M`.
- Mac mini: this branch's packaged Buzz running a saved
`buzz-agent`/relay-mesh agent.
- Confirmed `:9337` accepted inference and the saved ACP harness started
with no error while `:3131` was still unavailable.
- Exact inference succeeded before restart (`CORRECTED-MINI-E2E-OK`).
- Bespoke Buzz shut down cleanly in 2.3s, then restored ingress and the
saved harness in 18.8s while `:3131` was still unavailable.
- Exact inference succeeded after restart (`AFTER-RESTART-E2E-OK`).
- A real Buzz `@C55` message traversed desktop → `buzz-acp` →
`buzz-agent` → mini `:9337` → M5 compute and published the requested
reply successfully.
---------
Signed-off-by: Bartok9 <danielrpike9@gmail.com>
Signed-off-by: Michael Neale <michael.neale@gmail.com>
Co-authored-by: Bartok9 <danielrpike9@gmail.com>
Co-authored-by: Michael Neale <michael.neale@gmail.com>
## Problem
The Databricks model dropdown offers a handful of stale models — and
there's no way to tell that list apart from the real one. The AI Gateway
exposes **66** chat/embedding endpoints on `block-lakehouse-production`,
but the picker was showing a short list that includes models the gateway
no longer serves and embedding endpoints that can't chat at all.
Three independent defects, all on the discovery path:
**1. Live discovery never ran for agents with no saved provider.**
`get_agent_models` gates every in-process discovery attempt on the
provider (`is_openai_compatible_provider` / `is_anthropic_provider` /
`is_databricks_provider`), reading it straight from `record.provider`.
That field is `null` for every agent record created before provider
persistence — and for any agent that inherits its provider from the
build. So all three gates saw `None`, no HTTP discovery ran, and the
request fell through to the `buzz-acp models` subprocess. On the
Databricks path that subprocess returns `discovery_failure_fallback` —
the small hardcoded `DATABRICKS_V2_KNOWN_MODELS` catalog — which the
frontend renders exactly like a live catalog. An internal DMG that bakes
`BUZZ_AGENT_PROVIDER=databricks_v2` and a `DATABRICKS_HOST` still got
the fallback.
**2. The fallback list couldn't represent the running model.**
When discovery genuinely fails, the picker should at minimum be able to
show what the agent is actually configured with. For `DatabricksV2` it
couldn't: the fallback returned only the hardcoded slate, so a model
like `databricks-gpt-5-5` wasn't selectable in its own picker.
**3. Embedding endpoints were offered as chat models.**
`databricks-bge-large-en` was selectable (visible in the dialog today).
The v2 endpoints payload carries no `task` or `state` field, so there is
nothing to filter on but the name.
## Changes
- **`effective_discovery_provider`** (new,
`desktop/src-tauri/src/commands/agent_models_env.rs`) — an explicit
provider (saved record value, or the create/edit dialog's current form
value) still always wins; when there is none, discovery falls back to
the runtime's own provider env var (`GOOSE_PROVIDER`,
`BUZZ_AGENT_PROVIDER`, …) read off the merged env, which by that point
already carries the baked build floor and the process env. Wired into
both `get_agent_models` and `discover_agent_models`.
`SavedAgentModelDiscoveryConfig` now carries `provider_env_var` from
`known_acp_runtime`, so each runtime reads *its own* key rather than a
shared guess.
- The relay-mesh branches in `discover_agent_models` deliberately keep
using `input.provider`: those key off a deliberate provider selection,
never a baked default.
- **Asserted vs inferred matters for missing credentials.** The OpenAI
and Anthropic gates error on a missing API key, while the Databricks
gate falls through; an inferred provider hitting the first two would
have replaced a working subprocess catalog with `config:
ANTHROPIC_API_KEY required` (`export GOOSE_PROVIDER=anthropic` is
goose's documented way to pick a provider, and it keeps the key in its
own keyring). So `effective_discovery_provider` returns a
`DiscoveryProvider` that remembers how the value was resolved, and
`required_env` only reports a missing credential for an asserted
provider. A wrong guess declines and lets the subprocess answer.
- **`is_chat_capable_endpoint`** (new,
`crates/buzz-agent/src/catalog.rs`) — applied in
`parse_v2_endpoints_page`. Drops `*embedding*` and segment-matched `bge`
/ `gte` endpoints; keeps everything unrecognised (fail-open, so a new
model family is never hidden). Segment matching is why it's `split('-')`
and not `contains`: a substring check would swallow legitimate names.
- **`discovery_failure_fallback`** for `Provider::DatabricksV2` now
leads with the configured model (deduped against the known slate,
blank-tolerant), so a failed discovery still yields a picker that can
show the running model. The configured model is trimmed once up front —
`resolve_model` doesn't trim, so a padded `DATABRICKS_MODEL` used to
slip past the dedupe and appear twice.
- **`sort_v2_endpoints_newest_first`** (new, second commit) — the
catalog is now ordered newest-first on each endpoint's
`created_timestamp`, ties broken by name. Previously Buzz sorted
nothing, so the gateway's own order reached the picker: it pages in two
phases (Databricks-managed, then workspace-created — the page token
decodes to `{"phase":"user"}`), each alphabetical, which buried
`databricks-claude-opus-5` 8th behind five older Claude endpoints and
`goose-claude-opus-5` — the newest endpoint in the catalog — 55th of 63.
Sorting in `fetch_v2_models` means both discovery paths inherit it with
no wire or type changes, and the combobox filter preserves incoming
order. Endpoints with an absent or unparseable timestamp sort last
rather than first, so a wire-shape change degrades to "unordered at the
bottom" instead of "shuffled to the top".
- The name tiebreak is load-bearing: eleven managed endpoints share one
placeholder timestamp (`1699610000000`), so without it their relative
order would vary between runs. That placeholder is also not always
accurate — a few genuinely recent endpoints
(`databricks-kimi-k2-7-code`, `databricks-llama-4-maverick`) land at the
bottom with the 2023 batch. The gateway offers nothing better to sort
on.
- Env/provider lookup helpers moved out of `agent_models.rs` into
`agent_models_env.rs`. This keeps the command module under the file-size
limit **without ratcheting the override up** — the existing 1079 entry
is untouched (file is now 1066 lines).
## Verification
Live against `block-lakehouse-production`, release build:
```
BUZZ_ACP_AGENT_COMMAND=$PWD/target/release/buzz-agent \
BUZZ_AGENT_PROVIDER=databricks_v2 \
DATABRICKS_HOST=https://block-lakehouse-production.cloud.databricks.com \
DATABRICKS_MODEL=databricks-gpt-5-5 \
./target/release/buzz-acp models --json
```
- before: 66 endpoints, including `databricks-bge-large-en`,
`databricks-gte-large-en`, `databricks-qwen3-embedding-0-6b`
- after: **63** endpoints, `[.models[] | select(.id |
test("embedding|-bge-|-gte-"))]` → `[]`
Top of the list after the sort commit:
```
goose-claude-opus-5 2026-07-24
databricks-claude-opus-5 2026-07-23
databricks-gemini-3-6-flash 2026-07-20
databricks-gemini-3-5-flash-lite 2026-07-20
databricks-inkling 2026-07-14
```
Tests: 15 new (8 in `catalog.rs` — including the two-wire-shape
timestamp parse, the sort's tiebreak/no-timestamp cases, and the
padded-model dedupe — and 7 plus one assertion in
`agent_models_tests.rs`, 3 of them covering the asserted/inferred
credential split), two existing tests updated. `just check`, `just
test-unit`, and `just desktop-tauri-test` all pass (1636 desktop-tauri
tests, 274 buzz-agent lib tests).
Not run locally: the Docker-backed integration suite (`just test`) —
this diff touches neither `buzz-relay`, `buzz-db`, nor `buzz-auth`.
## Follow-ups (deliberately out of scope)
Two inference-path defects found while investigating, both reproduced
live against the gateway and both independent of discovery:
1. **Gemini thought signatures are dropped.** The gateway returns a bare
`thoughtSignature` on tool calls; the external-model serving endpoints
return it nested as `extra_content.google.thought_signature`. Neither
shape is round-tripped, so multi-turn tool use on `databricks-gemini-*`
fails with a 400 on the second turn.
2. **Array-shaped `content` is silently discarded.** Some models return
OpenAI `content` as a block array rather than a string; `parse_openai`'s
`str_field` returns `None` and the text is dropped.
The legacy `serving-endpoints` path does not work around either one, and
costs reasoning support on the GPT-5 family.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
## Why
The Codex adapter install-plan test depended on the process-global
login-shell PATH cache, making it flaky under concurrent or loaded CI
runs.
## What
- Thread an explicit probe PATH through adapter install planning
- Use a controlled PATH in Codex install-plan tests
- Update the existing desktop file-size allowance for the focused seam
## Risk Assessment
Low — production behavior keeps using the same augmented PATH; only
dependency injection and deterministic tests change.
## References
- Failure:
https://github.com/block/buzz/actions/runs/30110680608/job/89539202266
- Validation: `just desktop-tauri-test`; `cd desktop && pnpm check`;
full pre-push hooks
Generated with Codex
---------
Signed-off-by: npub1x4hk035p3p9q39a3fcrd2fe30lpkrhr5dwe0cqzzjphxyyh8m0gsq4vqap <356f67c681884a0897b14e06d527317fc361dc746bb2fc0042906e6212e7dbd1@buzz.block.builderlab.xyz>
Co-authored-by: npub1x4hk035p3p9q39a3fcrd2fe30lpkrhr5dwe0cqzzjphxyyh8m0gsq4vqap <356f67c681884a0897b14e06d527317fc361dc746bb2fc0042906e6212e7dbd1@buzz.block.builderlab.xyz>
Co-authored-by: Codex <noreply@openai.com>
## Summary
- Replace all references to a specific corporate VPN product name with
generic "VPN" / "corporate VPN" / "VPN tunnel" / "VPN CLI" language
across 15 files (21 lines)
- Comment-only and doc-only changes — zero functional impact
- Keeps the OSS repo free of vendor-specific assumptions
`rg -i warp` returns zero hits after this change (excluding
`node_modules`, lockfiles).
## Summary
- add a macOS local-network purpose string
- explain that LAN access is for optional Share Compute and local relay
connections
- clarify that messaging through a remote relay does not need it
## Context
macOS currently falls back to its generic Local Network prompt because
Buzz's `Info.plist` has no `NSLocalNetworkUsageDescription`. The alert
says Buzz can find and collect data from devices without explaining
which Buzz features use that capability, which makes an optional feature
feel invasive.
Buzz can intentionally make local-network connections for Share Compute
peers and user-configured local relays. Ordinary messaging through a
remote relay does not require that access.
Apple recommends that any app using the local network directly or
indirectly include `NSLocalNetworkUsageDescription`:
https://developer.apple.com/documentation/bundleresources/information-property-list/nslocalnetworkusagedescription
This PR improves the disclosure only. It does not change when macOS
asks. The exact unexpected trigger should be handled separately once it
can be reproduced and attributed.
## Verification
- `plutil -lint desktop/src-tauri/Info.plist`
- built a debug macOS `.app` bundle and confirmed its final `Info.plist`
contains the exact purpose string
- `just ci`
Signed-off-by: joelbrilliant <joelbrilliant1@gmail.com>
## Why
Installing the Codex, Claude, or Goose desktop app does not install the
command-line harness Buzz needs. The current UI makes that distinction
unclear, links some missing-CLI states to adapter documentation, and can
report a successful install from the installer exit code even when
runtime discovery still fails. On Windows, Buzz also invokes Goose's
Bash installer, which writes the executable somewhere Buzz does not
discover.
## What
- distinguish missing vendor CLIs from missing or outdated ACP adapters
in runtime metadata and UI guidance
- link Codex, Claude Code, and Goose missing-CLI states to their
official CLI installation documentation
- explain in Settings, onboarding, and agent configuration that the
desktop app alone is not sufficient
- use Goose's official PowerShell installer on Windows
- refresh PATH and rediscover the requested runtime after installation,
keeping the control retryable if the runtime is still unavailable
- add Rust and Playwright regression coverage for Windows installer
selection, CLI/adapter guidance, false-success prevention, verified
installs, and onboarding copy
## Risk Assessment
Medium. This changes desktop onboarding and runtime installation
behavior. Successful installs now require the runtime catalog to verify
availability; previously hidden discovery failures will surface as
actionable errors instead of a false success state.
## References
- [Codex CLI installation](https://developers.openai.com/codex/cli/)
- [Claude Code
installation](https://code.claude.com/docs/en/getting-started)
- [Goose
installation](https://goose-docs.ai/docs/getting-started/installation/)
- Follow-up to #2563 and #2587
## Validation
- `just desktop-typecheck`
- `just desktop-test` — 3,455 passed
- focused Rust post-install verification tests
- focused Playwright Doctor/onboarding coverage (in progress; CI and
local sequential rerun will provide final results)
Generated with Codex
---------
Signed-off-by: Atish Patel <atish@squareup.com>
Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: Goose <opensource@block.xyz>
## Problem
On Windows, `install_shell_command` wraps every install command in Git
Bash `-l -c`, including the Windows-specific `powershell.exe … irm
https://chatgpt.com/codex/install.ps1 | iex` command. A Git Bash login
shell prepends its POSIX dirs (`C:\Program Files\Git\usr\bin`) to PATH,
so when Codex's `install.ps1` shells out to bare `tar -xzf C:\…`, it
resolves Git's GNU tar (`/usr/bin/tar`) instead of Windows bundled
bsdtar. GNU tar parses `C:` as a remote host:
```
tar (child): Cannot connect to C: resolve failed
gzip: stdin: unexpected end of file
/usr/bin/tar: Child returned status 128
/usr/bin/tar: Error is not recoverable: exiting now
Downloaded Codex package archive did not contain the expected package layout.
```
Claude's installer doesn't hit the same failure because it doesn't shell
out to `tar`.
## Solution
On Windows, detect `powershell.exe` install commands and spawn them
natively (`Command::new("powershell.exe")`) instead of routing them
through Git Bash. The discriminator is a case-insensitive prefix check
on the first whitespace-delimited token — minimal and precise.
The native spawn preserves everything `install_shell_command` provides
that applies:
- `NPM_CONFIG_*` / `COREPACK` env strip + managed npm prefix env
- PATH composed from managed Buzz dirs + inherited process PATH (no
POSIX login-shell dirs)
- `CREATE_NO_WINDOW` so no console flash
- stdin null, piped-drain in `run_install_command`
- The retry/backoff/annotate logic is fully shared
The `-Command` body is split correctly at the boundary
(case-insensitive) and passed as a single argument to preserve pipes and
spaces inside the installer script call.
Non-PowerShell commands (e.g. `npm install -g …` adapter steps) continue
through the existing Git Bash path unchanged.
## Tests
6 unit tests:
1. `is_powershell_command` detection (positive + negative)
2. Routing: PowerShell → native spawn on Windows, non-PowerShell → Git
Bash
3. Unix: non-Windows path returns the shell command unchanged
(compile-time cfg)
4. `-Command` body preservation (no bash args in native spawn; body is
single arg)
Full `just desktop-tauri-test` suite: 1627/1627 passing. Windows CI will
validate end-to-end.
---------
Signed-off-by: Will Pfleger <pfleger.will@gmail.com>
Co-authored-by: npub1mn7jgtj4w2pd0g0zeuhxsa6jy6p0rewxz4kujt98my82ahfmp72sxjexk7 <dcfd242e557282d7a1e2cf2e6877522682f1e5c6156dc92ca7d90eaedd3b0f95@buzz.block.builderlab.xyz>
On Windows, the model dropdown never populates for CLI harnesses (Claude
Code, Codex) while the same flow works on macOS.
Model discovery spawns `buzz-acp models` with `PATH` taken from
`login_shell_path()` — which by design always returns `None` on Windows
(Git Bash's POSIX-shaped PATH would poison native children). The
discovery child therefore ran with only the raw inherited process PATH,
missing the Buzz-managed Node/npm directories and exe-parent sidecar
dir, so the ACP adapter's `.cmd` shims failed to resolve `node` and
discovery returned nothing. macOS worked only because a login-shell PATH
exists there.
The fix reuses the existing `augmented_path()` helper (already used by
CLI login probes and auth commands, built on the same
`build_augmented_path` kernel as the real agent spawn), so model
discovery resolves the identical toolchain the agent will actually run
with. `login_shell_path()` remains the login-shell component inside that
composition — on macOS the composed PATH is a superset of the previous
value.
Related: #2661 (managed Node fallback these entries point at), reported
in the Windows install-issues follow-up.
Signed-off-by: Will Pfleger <pfleger.will@gmail.com>
Co-authored-by: npub1fgdl5qqnh3k3f2xkqrvt7cujalhm623x4s7fdjdj5yrtp5fzjl9qrjpucw <4a1bfa0013bc6d14a8d600d8bf6392efefbd2a26ac3c96c9b2a106b0d12297ca@buzz.block.builderlab.xyz>
## Summary
- Preserve plain-string errors returned by Tauri and show an actionable
message when huddle audio is unavailable in the relay deployment.
- Use one error formatter across channel, timeline, wave, and profile
huddle actions while preserving other relay and device errors.
- Complete the huddle lifecycle when audio setup fails after publishing
a start event, preventing peers and reloaded clients from reconstructing
a phantom active huddle.
- Cover unavailable-audio formatting and START → rollback-END
reconstruction with regression tests.
## Behavior
The Tauri huddle commands reject with a plain string. Several desktop
toast call sites only preserved JavaScript `Error` objects, so the relay
message was discarded and replaced with “Failed to join huddle.”
The desktop now recognizes `huddle_audio_unavailable` and the current
relay message, then shows:
> Huddle audio isn’t available on this server. Ask an administrator to
turn it on.
Other relay and device messages remain intact, including microphone
errors.
`start_huddle` also publishes `KIND_HUDDLE_STARTED` before audio setup.
If setup fails, rollback now publishes `KIND_HUDDLE_ENDED` through the
normal end-and-archive path before resetting local state. This makes the
failed start observable to lifecycle reconstruction and prevents stale
join affordances.
## Checks
- `cd desktop && pnpm test` — 3,405 passed
- `cd desktop/src-tauri && cargo test` — 1,560 passed, 13 ignored; 3
diagnostic tests passed
- `cd desktop && pnpm exec playwright test tests/e2e/channels.spec.ts
--project=smoke --grep 'huddle rollback end event'` — passed
- `just desktop-tauri-check`
- `cd desktop && pnpm typecheck`
- `cd desktop && pnpm check`
- Pre-commit and pre-push hooks
Thomas Petersen