Commit Graph
6973 Commits
Author SHA1 Message Date
Slava Zavadsky b8779dae63 fix(coding-agent/eval): remove per-call model override from agent() (#6438)
Completes the maintainer's removal of per-call model selection from
subagent spawns (9f8aa87dbf removed it from the task tool and the
model-facing agent() docs/prompt, but the eval agent() runtime and all
four preludes still accepted and forwarded a per-call model).

Subagents now always resolve through the selected agent's frontmatter
model and settings, so an explicit model: "default" can no longer
silently route children onto the parent session model.

- agent-bridge: drops "model?" from agentArgsSchema and the request
  forward; adds "+": "delete" so a legacy model argument is stripped
  (same contract as the task wire schemas).
- JS/Python/Ruby/Julia preludes: remove the model parameter from
  agent(); completion()'s tier selector is unchanged.
- docs (tools/eval.md, python-repl.md) updated to the removed surface.

Refs #6438
2026-08-04 08:35:01 -04:00
can1357 ad37de9663 Merge PR #7553: fix(coding-agent): allow quoted metacharacters in bash patterns (@roboomp) 2026-08-04 01:03:01 +02:00
can1357 fbfae3b4ad Merge PR #7540: fix(discovery): honor disabled codex mcp servers (@roboomp) 2026-08-04 01:02:56 +02:00
roboomp 98a65ffbdf fix(coding-agent): blocked escaped reinterpreted payloads
- Preserved escaped control characters for the downstream reinterpretation safety decision.
- Covered the backslash-escaped git inline shell-alias bypass.

Fixes #7552
2026-08-03 21:18:35 +00:00
roboomp b621a9a637 fix(coding-agent): flagged double-quoted reinterpreted payloads
- Treated double-quoted shell-control chars like single-quoted ones so a -c/-e reinterpretation option still gates them.
- Covered the double-quoted git inline shell-alias bypass.

Fixes #7552
2026-08-03 21:13:13 +00:00
roboomp 9bcd31fcd2 fix(coding-agent): blocked reinterpreted quoted shell payloads
- Kept quoted shell controls prompt-gated when code-evaluation options can reinterpret the argument.
- Covered the reported git inline shell-alias bypass while retaining Cargo regex approval.

Fixes #7552
2026-08-03 21:08:52 +00:00
roboomp 54b2e38564 fix(coding-agent): allowed quoted bash pattern metacharacters
- Replaced the raw character guard with quote-aware scanning while retaining command substitution and unquoted shell-control protections.
- Added regression coverage for the reported Cargo benchmark filter.

Fixes #7552
2026-08-03 20:54:41 +00:00
can1357 bc39ffa265 feat: introduced omptype validation package and migrated workspace dependencies
- Introduce `@oh-my-pi/omptype` as a new ArkType-compatible schema validation package featuring a lazy JIT runtime, JSON Schema emission, and compatibility adapters.
- Replace `arktype` across workspace packages and test utilities with `@oh-my-pi/omptype`.
- Add benchmark suites, tests, and documentation for the new validation engine and adapters.
- Update workspace build, test runner, and release configurations to include the new package.
2026-08-03 21:56:48 +02:00
can1357 b48aeacf25 fix(coding-agent/eval): made cells own every bridge call they start
- Split the Python bridge signal: the raw abort reaches tools (so
  subagents die with the turn) while a shielded signal governs how long
  the host waits, so a cancel can no longer settle a cell on top of a
  still-running isolation merge.
- Mirrored the contract in the JS runtime: abort in-flight tool calls
  immediately, then drain any deferExternalAbort phase before killing
  the worker, and refuse new bridge calls once cancelled.
- Held a finished worker result until the run's tool calls drain. A
  floated agent() previously settled the cell at once, dropping the
  run's abort listener and leaving the subagent running with nothing
  able to cancel it.
- Added regression coverage for all three, each verified to fail
  without its fix.
2026-08-03 19:55:20 +02:00
can1357 c7a3010d7c fix(coding-agent/eval): ensured tool bridge receives unshielded abort signal
- Prevent eval agent bridge calls from inheriting the kernel abort shield.
- Unset maxRuntimeMs in runEvalAgent to properly inherit task.maxRuntimeMs.
2026-08-03 18:57:22 +02:00
roboomp 6f7600cd89 fix(discovery): prioritized project codex mcp entries
Load project Codex MCP entries before user entries so a disabled project server claims its dedupe key before a same-named user server can survive.

Added regression coverage for project-over-user disable precedence.

Fixes #7538
2026-08-03 16:52:54 +00:00
roboomp 012836d99e fix(discovery): tag disabled codex mcp servers instead of dropping
Carry enabled = false through discovery so loadAllMCPConfigs' suppress
path can claim the dedupe key (keeping a same-named lower-priority
source disabled) and honor the user force-enable allowlist. Dropping the
entry outright defeated both.

Fixes #7538
2026-08-03 16:45:31 +00:00
roboomp 8b854598f1 fix(discovery): honored disabled codex mcp servers
- Skipped Codex config.toml MCP entries explicitly marked enabled = false.
- Added discovery regression coverage for disabled and enabled entries.

Fixes #7538
2026-08-03 16:38:20 +00:00
can1357 1a8caad23e chore: update docs + rename reset to clear 2026-08-03 18:37:23 +02:00
roboompandcan1357 f948c61256 fix(discovery): enforced hard model probe timeouts
- Rejected local model discovery at the configured deadline even when fetch ignored abort.
- Covered pending-transport behavior with deterministic fake timers.

Fixes #7482
2026-08-03 16:43:52 +02:00
can1357 4ddc2d2cd4 chore: rewrite changelogs + fix stale tests 2026-08-03 15:32:19 +02:00
can1357 39bc9de52f style: applied biome import order and formatting 2026-08-03 15:26:20 +02:00
can1357 9fb082bf14 refactor(utils): consolidated file locking into pi-utils file-lock
- Moved the coding-agent lock-directory primitive to @oh-my-pi/pi-utils/file-lock
  and migrated settings, MCP config-writer, and security store imports.
- Replaced the stats aggregator's parallel ~200-line token/breaker lock protocol
  with the shared primitive: dead owners reclaimed immediately, live-but-wedged
  owners after STATS_SYNC_LOCK_STALE_MS, unstamped acquisitions after the new
  acquireStaleMs grace (10s).
- Shared primitive now treats EPERM kill probes as live owners.
- Rewrote the stats lock-reclamation regressions against the shared protocol
  and moved the file-lock contract test into pi-utils.
2026-08-03 15:25:03 +02:00
can1357 c53a47f97d Merge PR #7287: fix(coding-agent): prune archived session stats (@alphastorm)
# Conflicts:
#	packages/coding-agent/src/session/session-manager.ts
2026-08-03 15:15:36 +02:00
can1357 267856deb8 Merge PR #7475: fix(session): prevent stale /btw branch promotion (@roboomp) 2026-08-03 15:12:03 +02:00
can1357 6a44844c57 Merge PR #7377: fix(hub): wake owners when supervised processes exit (@paralin) 2026-08-03 15:12:03 +02:00
can1357 1b5148ed84 Merge PR #7368: fix(browser): guard cmux guest rejections (@roboomp) 2026-08-03 14:46:24 +02:00
can1357 b1d75bb55b fix(session): preserve title on first-turn branches 2026-08-03 14:46:24 +02:00
can1357 14a03d4579 Merge PR #7477: fix(session): preserve title when branching (@lemonleks) 2026-08-03 14:46:24 +02:00
can1357 d46c58b5c8 fix(tui): clear extension command image drafts 2026-08-03 14:46:24 +02:00
Aleksandr Khaustov 276ff92afe fix(tui): consume extension slash commands locally 2026-08-03 14:02:05 +04:00
roboomp a54a2f79cb fix(tui): preserved b before /btw completion
Reserve the plain b shortcut only after /btw has a completed answer or a branch is already pending. Running, empty, aborted, and failed panels now leave the key for the composer, while completed-but-refused branches still consume it with an explanation.

Fixes #7474
2026-08-03 09:33:07 +00:00
Aleksandr Khaustov efe640a161 fix(session): preserve title when branching 2026-08-03 13:27:01 +04:00
roboomp 98f484bd2f fix(session): validated session identity for /btw branch
Branched session files preserve entry ids, so leaf-id equality alone let a stale /btw answer promote into a different loaded session. Capture the originating session id at /btw start and require it to match at both the controller gate and every branchFromBtw checkpoint.

Fixes #7474
2026-08-03 09:22:32 +00:00
roboomp 7a1183da91 fix(session): prevented stale /btw branch promotion
- Passed the authorized leaf through the branch executor and revalidated it before rewriting history.
- Refused promotion during active turns and bounded post-prompt drains.
- Consumed unavailable branch keys while showing pending and refusal state in the panel.

Fixes #7474
2026-08-03 09:07:12 +00:00
Christian Stewart 302148523f fix(hub): wake owners when supervised processes exit
Publish terminal daemon completions to the session that started the
process so idle agents can resume without polling hub status.

Persist every unacknowledged generation with a stable completion ID and
immutable snapshot. Replay the collection after reconnect or broker
recovery, and clear each event only after the owning client acknowledges
it.

Signed-off-by: Christian Stewart <christian@aperture.us>
2026-08-03 01:31:39 -07:00
can1357 f41f3f3ee1 test(read): deraced shared PDF extraction abort tests
- Aborting a caller while it was the sole extraction waiter tore the shared
  extraction down and deadlocked against the blocked conversion mock, hanging
  the CI chunk until SIGKILL.
- Sequenced owner/joiner starts and added an untilAborted spy barrier that
  waits for both waiters to attach before aborting.
2026-08-03 06:43:54 +02:00
can1357 63b07f8ec0 chore: rewrite changelogs 2026-08-03 05:52:53 +02:00
can1357 5ad263f0a6 fix(tui): hid stripped-tool-call placeholder with tool activity
- The 'N tool calls elided' replay placeholder leaked tool activity while
  display.hideToolActivity was on; it is now a visibility-aware component
  wired into both the hotkey and /settings toggle paths.
- Added replay + live-reveal regression coverage.
2026-08-03 05:42:07 +02:00
can1357 0244e18e43 fix(extensions): refreshed formerly synchronous sources on reload walks
- A reload that drops a module's last require() edge leaves the permanent
  hooks serving it from the synchronous snapshot map, which was only
  refreshed while the path stayed flagged; an edit after the downgrade
  replayed stale bytes. Ensure now re-rewrites and refreshes the snapshot
  for every ever-synchronous path on each graph walk.
- Added the mirror reload regression (require edge dropped + source edited).
2026-08-03 05:42:07 +02:00
can1357 03c07f89ec Merge PR #7342: feat(tui): add hidden tool activity mode (@dannyboy-ai) 2026-08-03 05:34:17 +02:00
can1357 e1ac59e5df fix(extensions): served sync-upgraded modules from earlier async load hooks
- A reload that adds a require() edge to an already-hooked ESM module never
  re-registers hooks, and the original async onLoad filter keeps matching;
  require() rejects async onLoad results, so the async hook now serves the
  pre-rewritten synchronous source inline when one exists.
- Added a same-process reload regression covering the async-to-sync upgrade.
2026-08-03 05:33:06 +02:00
can1357 c83bff390d Merge PR #7406: fix(coding-agent): restore legacy compaction exports (@roboomp) 2026-08-03 05:26:10 +02:00
can1357 ed7ea77a6b Merge PR #7405: fix(extensions): resolve compiled dependency graphs (@roboomp) 2026-08-03 05:26:10 +02:00
Daniel Anderson-Little 4471a3f26d fix(tui): purge hidden tool images 2026-08-02 23:21:17 -04:00
roboompandcan1357 8f7c7f7415 fix(coding-agent): honor vision role thinking effort in inspect_image
inspect_image resolved @vision with resolveModelFromString, which dropped the
:high thinking selector, and passed no reasoning to the oneshot. The
google-gemini-cli mapper then emitted thinkingBudget: 0, which thinking-only
Gemini models reject with HTTP 400. Resolve the role's explicit thinking
selector, clamp it to the model's supported efforts, and forward it as the
oneshot reasoning.

Fixes #7448
2026-08-03 05:16:14 +02:00
can1357 4b8c19366d Merge PR #7437: fix(edit): prevent fuzzy replace-all self-matching (@roboomp) 2026-08-03 05:15:03 +02:00
can1357 a95476bad6 Merge PR #7441: fix(mcp): expand env vars in reauth OAuth credentials and reject empty tokens (@roboomp) 2026-08-03 05:15:03 +02:00
can1357 b8c81e195f Merge PR #7444: fix(coding-agent): keep completed session entries durable across crashes (@olegpulatov) 2026-08-03 05:15:02 +02:00
can1357 a58584ff3c feat: implemented shared serialization utilities and standardized inventories
- Added shared Python call and literal serialization utilities with multiline verbatim support.
- Standardized tool inventories to format as an OpenAI-Harmony functions namespace using TypeScript declarations.
- Updated tool normalization and rendering functions to accept options objects and default to Python-syntax examples.
- Refactored Gemini dialect rendering to leverage shared serialization functions directly.
2026-08-03 05:06:49 +02:00
Daniel Anderson-Little a20690a40b feat(tui): add hidden tool activity mode 2026-08-02 22:59:48 -04:00
Oleg Pulatov 6c84a8bb99 fix(coding-agent): keep completed session entries durable across crashes
Completed transcript entries now write through to the OS page cache on
append instead of microtask-batching, supersede in-flight atomic rewrites
with a synchronous full-body publish, and land on the live moveTo path
(source pre-rename, destination post-rename) so a software crash no longer
drops finished user/assistant/tool events. Streaming text remains durable
only at message_end; no fsync/power-loss guarantee is claimed.
2026-08-03 04:14:58 +02:00
roboomp 4d74fabe70 test(mcp): restored OAuth env after reauth coverage
Preserved pre-existing MCP_OAUTH_CLIENT_ID and MCP_OAUTH_CLIENT_SECRET values
instead of deleting them after the env-expansion regression test, preventing
later tests in the same Bun process from observing mutated caller state.

Fixes #7440
2026-08-03 00:37:55 +00:00
roboomp c6a057073b fix(mcp): expand env vars in reauth oauth credentials and reject empty tokens
/mcp reauth read OAuth clientId/clientSecret from the raw, unexpanded config
while URL and resource used expandEnvVarsDeep, so `${VAR}` placeholders were
sent literally to the token exchange. MCPOAuthFlow.exchangeToken() also accepted
any HTTP-success body, storing an empty access token when a provider signals
failure with HTTP 200 (e.g. Slack `{ ok: false, error }`), surfacing only later
as invalid_token.

- Select flow client credentials from runtimeBaseConfig / expanded auth block;
  keep the raw placeholder for the persisted config file.
- Reject token responses without a non-empty access_token, including the
  sanitized provider error when present.
- Add regression tests for env-expanded reauth credentials and HTTP-200 token
  error bodies.

Fixes #7440
2026-08-03 00:33:36 +00:00
roboomp aad12ab2f7 fix(edit): prevented fuzzy replace-all self-matching
Matched fuzzy candidates against immutable source text while excluding ranges already selected for replacement. Inserted replacement content can no longer become a later exact or fuzzy candidate.

Added regression coverage for multiple fuzzy source matches when the replacement contains the original search text.

Fixes #7432
2026-08-03 00:00:58 +00:00