Commit graph

54 commits

Author SHA1 Message Date
4a29c13e0d feat: steer active chat messages
Co-authored-by: Codex <noreply@openai.com>
2026-03-09 16:56:24 -07:00
Hari
a018a0cd3e Merge pull request #281 from getcompanion-ai/codex5.3
model upgrade
2026-03-09 19:38:00 -04:00
a1c66c805a model upgrade 2026-03-09 16:37:39 -07:00
Hari
92c3bcf9f9 Merge pull request #277 from getcompanion-ai/fix-package
fix loading
2026-03-09 17:42:51 -04:00
cc7f68d4d2 fix loading 2026-03-09 14:37:59 -07:00
Hari
a46e9ede61 Merge pull request #276 from getcompanion-ai/fix-package
fix grind package installation
2026-03-09 16:47:39 -04:00
0fbb44b575 fix grind package installation 2026-03-09 13:43:14 -07:00
Hari
7f72ee175d Merge pull request #275 from getcompanion-ai/prompt
prompt
2026-03-09 16:33:06 -04:00
ffb6f1db70 prompt 2026-03-09 13:31:12 -07:00
Hari
d880d2740e Merge pull request #273 from getcompanion-ai/grind
fix chat
2026-03-09 16:02:14 -04:00
Hari
064517f749 Merge pull request #274 from getcompanion-ai/prompt
update
2026-03-09 15:53:01 -04:00
3c0f74c1dc fix(coding-agent): harden chat stream completion
Flush final text before closing each AI SDK text block, surface event-processing failures to chat callers, and clear the remaining Companion OS check blockers.

fixes #273

Co-authored-by: Codex <noreply@openai.com>
2026-03-09 12:48:21 -07:00
85c2a0b675 rm bootstrap 2026-03-09 12:44:17 -07:00
bf9bfe38f7 update 2026-03-09 12:34:45 -07:00
6b2a639fb6 fix chat 2026-03-09 10:46:43 -07:00
43c6b56dfa fix chat 2026-03-09 10:01:59 -07:00
Hari
f58ff785ce Merge pull request #272 from getcompanion-ai/grind
Add durable grind mode and non-mutating Companion OS checks
2026-03-09 12:26:52 -04:00
48fa514e7e fix(companion-os): address grind mode review feedback
Restore the bedrock provider import through an explicit workspace path mapping, align pi-grind package metadata with the existing package manifests, remove the unused controller-failure state path, and parse bare ISO dates in local time.

Co-authored-by: Codex <noreply@openai.com>
2026-03-09 00:02:00 -07:00
8c92f479d6 chore(companion-os): make checks non-mutating
Switch the apps/companion-os check script from biome check --write to biome lint so verification stops rewriting the worktree while still running tsgo and the browser smoke bundle.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 23:50:01 -07:00
ff6e39dd10 grind mode baby 2026-03-08 23:38:13 -07:00
Hari
4c976822c4 Merge pull request #269 from getcompanion-ai/prompt
prompt
2026-03-09 01:54:48 -04:00
Hari
444ab5820d Merge pull request #270 from getcompanion-ai/chat-single
single chat sot
2026-03-09 01:54:04 -04:00
bc68e30b6a personality 2026-03-08 22:25:57 -07:00
e4ee3e64f4 single chat sot 2026-03-08 22:12:03 -07:00
5308b63671 rm heartbeat 2026-03-08 21:19:30 -07:00
aa70afbd7e align skills to pi native 2026-03-08 20:14:24 -07:00
59ad12335a Align pi sandbox context and bootstrap injection
Co-authored-by: Codex <noreply@openai.com>
2026-03-08 17:41:57 -07:00
Hari
fb782fa025 Merge pull request #262 from getcompanion-ai/memory-staging
Add first-class memory management plumbing
2026-03-08 19:54:01 -04:00
9e11f49d17 Fix runtime memory review issues
Address runtime memory review feedback around sqlite compatibility, shutdown ordering, and endpoint validation.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 16:20:56 -07:00
9c5dde8615 Drop stale pi-memory-md lockfile entry
Remove the orphaned pi-memory-md workspace stanza after deleting the unused package directory.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 15:52:46 -07:00
bba47c28c4 rm 2026-03-08 15:51:12 -07:00
5c389efcf9 Make memory runtime-native for companion chat
Replace the old project-scoped file memory plumbing with runtime-native conversational memory and remove obsolete pi-memory-md shipping/wiring.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 15:24:52 -07:00
Hari
63e33460f6 Merge pull request #260 from getcompanion-ai/codex/deterministic-hosting
deterministic hosting
2026-03-08 17:59:58 -04:00
d0835a13d6 merge: bring staging into deterministic hosting
Co-authored-by: Codex <noreply@openai.com>
2026-03-08 14:55:17 -07:00
f0e5e26aac fix 2026-03-08 14:42:54 -07:00
5fe8f493b9 fix: address tmux adapter review feedback
Co-authored-by: Codex <noreply@openai.com>
2026-03-08 14:23:11 -07:00
fb88f43573 teams tmux 2026-03-08 14:11:48 -07:00
Hari
9057c30726 Merge pull request #263 from getcompanion-ai/codex/browser-first-class-agent-browser
feat: add built-in browser tool
2026-03-08 16:23:16 -04:00
Hari
c9c458f9cc Update apps/companion-os/packages/coding-agent/src/cli/args.ts
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
2026-03-08 16:22:41 -04:00
Hari
cca06e2235 Update apps/companion-os/packages/coding-agent/src/cli/args.ts
Co-authored-by: greptile-apps[bot] <165735046+greptile-apps[bot]@users.noreply.github.com>
2026-03-08 16:22:32 -04:00
9765576c0a fix: report unavailable memory repo status
Treat git status failures as unavailable instead of clean in gateway memory status.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 12:57:36 -07:00
4f2dd90d0f fix: address browser review feedback
Handle null browser exits, preserve empty-string wait targets, and avoid creating browser directories before action validation.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 12:57:22 -07:00
e0801dedb0 fix: address memory review findings
Remove shell-based memory tree inspection and pre-sync before memory pushes.
Switch sync update detection to git HEAD comparisons instead of localized output parsing.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 12:56:39 -07:00
0720c47495 feat: add built-in browser tool
Add a first-class browser tool backed by agent-browser and enable it by default in coding sessions.

Include CLI and system-prompt wiring plus focused coverage for the new tool.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 12:50:50 -07:00
2886855706 feat: add first-class memory management
Expose gateway memory APIs for status, init, files, search, and sync.
Align pi-memory-md with project-scoped, local-first memory behavior.

Co-authored-by: Codex <noreply@openai.com>
2026-03-08 12:47:38 -07:00
Hari
df702d95a3 Merge pull request #254 from getcompanion-ai/fix/bash-default-timeout
fix: add default timeout and no-output timeout to bash tool
2026-03-07 23:13:24 -05:00
973baf58e1 refactor: clean up no-output timeout architecture
- Remove onKillHandle from BashOperations interface (implementation detail leak)
- Use AbortController + combineAbortSignals to kill from tool level instead
- Extract shared buildOutput() to deduplicate .then()/.catch() paths
- Cleaner separation: BashOperations stays unchanged, timeouts are tool-level

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-07 19:58:49 -08:00
479124d945 fix: add default timeout and no-output timeout to bash tool
Bash tool had no default timeout, so commands like `npx serve` would
block the agent loop forever. This caused the randomtrends sandbox
thread to die - the LLM started a dev server, the shell never exited,
and the entire session hung with no way to recover.

Two safeguards:
- Default 120s timeout when LLM does not specify one
- 30s no-output timeout that kills processes that go silent

Both are configurable via BashToolOptions and can be disabled by
setting to 0. The LLM can still pass explicit higher timeouts for
long builds.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-07 19:28:50 -08:00
Hari
19f11932e4 Merge pull request #253 from getcompanion-ai/codex-chat-durable-gateway-recovery
feat: persist chat threads and recover gateway sessions
2026-03-07 20:19:08 -05:00
357ec23c6b chore: remove unused pi web-ui package
Delete the dormant companion-os web-ui workspace and clean up the workspace, build, lint, and lockfile references that still pointed at it.

Fixes #253

Co-authored-by: Codex <noreply@openai.com>
2026-03-07 17:12:20 -08:00