sanil-23 and GitHub
c818f026cb
fix(mcp): surface 401s as an actionable "needs sign-in" state ( #3719 )
...
Closes #3719 . MCP HTTP 401s now report a distinct `Unauthorized` state (typed McpUnauthorizedError + atomic ConnectFailure snapshot) with a localized 'Sign in' notice instead of a raw-error dead-end; recovery polling + i18n across 14 locales.
2026-06-17 13:11:28 +05:30
sanil-23 and GitHub
cdadf316aa
fix(chat): only deliver proactive messages to a fresh thread ( #3713 )
...
Gate proactive (morning brief / subconscious) delivery on thread freshness via threadHasMessages(); reuse a thread only when provably empty, else open a new thread. Unknown/unloaded metadata fails closed. Closes #3713 .
2026-06-17 13:10:20 +05:30
sanil-23 and GitHub
9abeda33c3
fix(scheduler-agent): scope boundary so calendar/meet-link reads don't mis-route to the scheduler ( #3717 ) ( #3731 )
2026-06-17 13:10:09 +05:30
sanil-23 and GitHub
bd5db5e95f
fix(conversations): render task insights panel once after full response ( #3717 Bug 2) ( #3726 )
2026-06-17 13:09:31 +05:30
Zhang and GitHub
f3086ca90c
test(observability): cover session-expired before_send events ( #3662 )
2026-06-16 18:43:49 +05:30
Zhang and GitHub
eae73c9de0
test(observability): cover updater GitHub 403 message events ( #3661 )
2026-06-16 17:07:26 +05:30
Zhang and GitHub
171b5ee140
test(observability): cover custom_openai Ollama timeout chain ( #3660 )
2026-06-16 17:06:59 +05:30
Zhang and GitHub
7dbbb9d71c
test(observability): cover custom_openai 502 transient shape ( #3659 )
2026-06-16 17:06:03 +05:30
Zhang and GitHub
34fb879f71
test(observability): cover Composio list_connections 503 wrappers ( #3658 )
2026-06-16 17:05:05 +05:30
afab6e5ede
fix(embeddings): prevent + handle LM Studio "No models loaded" embed flood (Sentry TAURI-RUST-4P4) ( #3687 ) ( #3688 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-16 00:19:42 -07:00
461f39bd6b
feat(agent-team): live teammate execution + messaging ( #3374 ) ( #3637 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-16 12:49:31 +05:30
6b6db6129a
feat(memory): windowed cover_window tool + 24h morning brief ( #3700 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 23:39:28 -07:00
756b434821
feat(agent): vision-v1 tier + vision sub-agent for image understanding ( #3699 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 19:53:37 -07:00
1e0985f576
[merge 3/5] test: agent harness behaviors JSON-RPC E2E suite ( #3471 ) ( #3667 )
...
Co-authored-by: Claude Fable 5 <noreply@anthropic.com >
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 19:41:43 -07:00
bf6fd85006
feat(intelligence): worktree isolation frontend + RPC wiring ( #3376 ) ( #3696 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 19:41:26 -07:00
6c89396e6f
fix(inference): demote BYO provider 401/403 auth failures from Sentry ( #3671 ) ( #3689 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 19:40:47 -07:00
7811bb0ab8
feat(agentbox): GMI Cloud AgentBox marketplace adapter ( #3620 ) ( #3651 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 19:02:51 -07:00
a2051b1307
fix(inference): use Codex client version for OAuth ( #3414 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 18:32:07 -07:00
oxoxDev and GitHub
ae24451b2a
feat(intelligence): workflow orchestration UI + approval card ( #3375 ) ( #3697 )
2026-06-15 18:29:12 -07:00
976b31f631
refactor(cdp): migrate all scanners to in-process transport; drop TCP DevTools port ( #3636 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 18:27:55 -07:00
445b11afc2
fix(inference): cap extraction max_tokens + classify provider 402 to stop the retry/Sentry storm ( #3616 ) ( #3617 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 18:27:39 -07:00
5b80f25cd4
[merge 1/5] feat(approval): debug-only OPENHUMAN_APPROVAL_TTL_SECS TTL override ( #3471 ) ( #3665 )
...
Co-authored-by: Claude Fable 5 <noreply@anthropic.com >
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 18:26:54 -07:00
363290fcea
feat(agent): LLM auto-compact summarization for subagents ( #3694 )
...
Co-authored-by: Claude <noreply@anthropic.com >
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 18:26:39 -07:00
93c7b3d6a3
fix(agent): parse Claude-native <invoke name> attribute tool calls ( #3622 )
...
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com >
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 17:20:08 -07:00
0adbed247d
perf(memory): swap-not-clone the chunk body in extract scoring ( #3436 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 17:16:59 -07:00
mysma-9403 and GitHub
3ae10f19c8
perf(notion): bound-concurrent page body markdown fetch ( #3433 )
2026-06-15 17:15:49 -07:00
a892cd4206
perf(memory-sources): batch composio reconcile upserts into one config save ( #3432 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 17:15:33 -07:00
mysma-9403 and GitHub
8709c53ebf
perf(team): cache managed-tool budget probe behind a short TTL ( #3431 )
2026-06-15 17:14:14 -07:00
mysma-9403 and GitHub
7802d1a618
perf(mcp_registry): spawn installed MCP servers concurrently at boot ( #3429 )
2026-06-15 17:13:33 -07:00
Zhang and GitHub
ce73f0e704
fix(inference): demote expected model test failures ( #3676 )
2026-06-15 17:05:13 -07:00
a32f361b15
[codex] fix numeric grounding in agent prompts ( #3678 )
...
Co-authored-by: Sami Rusani <sr@samirusani>
2026-06-15 17:04:26 -07:00
eb6f563d84
feat(telegram): transcribe incoming voice messages ( #3679 )
...
Co-authored-by: Ron <ron@SpinMaster.local >
2026-06-15 17:03:05 -07:00
7d71b06711
[merge 5/5] docs: agent harness E2E coverage matrix + plan ( #3471 ) ( #3668 )
...
Co-authored-by: Claude Fable 5 <noreply@anthropic.com >
2026-06-15 17:01:37 -07:00
a54854af03
[merge 4/5] test(e2e): agent harness behaviors browser spec ( #3471 ) ( #3669 )
...
Co-authored-by: Claude Fable 5 <noreply@anthropic.com >
2026-06-15 17:01:32 -07:00
6363790b1f
[merge 2/5] feat(channels/web): per-runtime approval-surface bridge for E2E tests ( #3471 ) ( #3666 )
...
Co-authored-by: Claude Fable 5 <noreply@anthropic.com >
2026-06-15 17:00:37 -07:00
2202835d40
Add memory_tree_delete_source RPC ( #3664 )
...
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com >
2026-06-15 17:00:00 -07:00
CodeGhost21 and GitHub
6b0a01684f
fix(embeddings): refuse empty embed inputs at provider + caller boundaries ( #13021 ) ( #3670 )
2026-06-15 16:58:46 -07:00
d89a7e8782
fix(deps): bump postgres stack to clear RUSTSEC-2026-0178/0179/0180 ( #3690 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 16:57:35 -07:00
CodeGhost21 and GitHub
d9591f73eb
fix(observability): skip SQLITE_FULL "database or disk is full" (Sentry TAURI-RUST-B6N) ( #3672 )
2026-06-15 16:32:09 -07:00
f5823be0ae
feat(settings): add Kimi API key affiliate link ( #3685 )
...
Co-authored-by: Huyuochi <07611331+Huyuochi@users.noreply.github.com >
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 16:29:11 -07:00
oxoxDev and GitHub
6ca8a5f33c
feat(command_center): add stop/retry/continue/follow_up control verbs ( #3373 ) ( #3698 )
2026-06-15 16:11:30 -07:00
eb80d249c6
fix(inference): use normalized role for Responses content kind ( #3624 ) ( #3626 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 13:29:34 -07:00
e4666f9be4
fix(memory): clear structure-degraded latch on any completed extraction ( #3365 ) ( #3631 )
...
Co-authored-by: Claude <noreply@anthropic.com >
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-15 13:28:23 -07:00
c5087d83da
feat(meet): in-call meeting agency (active replies, streaming, approvals) ( #3677 )
...
Co-authored-by: Claude Fable 5 <noreply@anthropic.com >
2026-06-15 11:34:01 -07:00
27739ee59e
fix(embeddings): prevent + handle custom endpoint with no embeddings API (Sentry TAURI-RUST-5JR) ( #3625 ) ( #3629 )
...
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-14 22:06:06 -07:00
2944b83fc0
fix: hardening audit pass 1 — core RPC auth, inference resilience, secrets, deep-link CSRF ( #3652 )
...
Co-authored-by: WOZCODE <contact@withwoz.com >
2026-06-14 12:26:05 -07:00
Steven Enamakel's Droid and GitHub
6ceee181d1
test(e2e): repair desktop Appium suite after IA refactors ( #3649 )
2026-06-13 13:28:39 -07:00
595fb41c9b
fix(memory): route only unrecoverable job failures to error status ( #3365 ) ( #3630 )
...
Co-authored-by: Claude <noreply@anthropic.com >
Co-authored-by: Steven Enamakel <enamakel@tinyhumans.ai >
2026-06-13 13:28:15 -07:00
sanil-23 and GitHub
e105e63c0c
feat(claude-code): opt-in agent provider + OpenHuman memory over MCP, macOS auth fix ( #3310 ) ( #3647 )
2026-06-14 01:40:52 +05:30
sanil-23 and GitHub
110fb5b08c
feat(mcp): MCP servers usable by the agent end-to-end — auth, OAuth, agent bridge ( #3495 ) ( #3648 )
2026-06-13 12:48:36 -07:00