Peter Steinberger
dbe156e881
fix(agents): sanitize transcripts for strict tool APIs
2026-01-11 04:46:18 +00:00
Peter Steinberger
f00038b383
fix(testing): stabilize live model runs
2026-01-11 04:22:35 +00:00
Peter Steinberger
9046296ed3
fix: clarify sub-agent sandbox limits
2026-01-11 05:04:14 +01:00
Peter Steinberger
30348e41c6
test: stabilize doctor + sandbox tests
2026-01-11 04:45:04 +01:00
Peter Steinberger
7343597075
chore: keep gate green
2026-01-11 04:42:44 +01:00
Peter Steinberger
50e62122bb
chore: format sandbox skills test
2026-01-11 04:39:42 +01:00
Peter Steinberger
57e6a9a762
fix: clamp z.ai developer role
2026-01-11 04:28:30 +01:00
Peter Steinberger
7660a78330
fix: mirror skills for read-only sandbox
2026-01-11 04:24:19 +01:00
Peter Steinberger
1f9b4e3af6
fix: send heartbeat reasoning ( #694 ) (thanks @antons)
2026-01-11 03:51:51 +01:00
Peter Steinberger
2b07a2a8ab
fix: stabilize onboarding auth tests ( #703 ) (thanks @mteam88)
2026-01-11 03:42:27 +01:00
Peter Steinberger
20b4e2b859
fix: stabilize live probes and docs
2026-01-11 02:26:39 +00:00
Peter Steinberger
6668805aca
fix(agents): enforce single-writer session files
2026-01-11 02:25:45 +00:00
Peter Steinberger
3a113b7752
fix: stabilize cli runner output
2026-01-11 02:25:45 +00:00
Peter Steinberger
4759633df1
fix(cli): keep build/lint green
2026-01-11 02:44:24 +01:00
Peter Steinberger
e824b3514b
fix(status): improve diagnostics and output
2026-01-11 02:42:24 +01:00
Peter Steinberger
02270abc87
feat: add codex cli backend
2026-01-11 01:39:30 +00:00
Peter Steinberger
a32021dc3e
fix: inject image paths for cli backends
2026-01-11 00:55:22 +00:00
Peter Steinberger
24c3ab6fe0
fix: unblock claude-cli live runs
2026-01-11 00:55:22 +00:00
Peter Steinberger
d8f1124d59
feat: add CLI backend fallback
2026-01-11 00:55:22 +00:00
Peter Steinberger
07be761779
feat: add sandbox browser control allowlists
2026-01-11 01:52:32 +01:00
Peter Steinberger
b339097179
style: format browser tool wiring
2026-01-11 01:34:45 +01:00
Peter Steinberger
326fb04d12
feat: add browser target selection for sandboxed agents
2026-01-11 01:31:56 +01:00
Shadow
c731a87d07
Discord: add fetch message action
2026-01-10 16:38:20 -06:00
Peter Steinberger
4d0e74ab6c
fix: cover extra thinking tags ( #688 ) (thanks @theglove44)
2026-01-10 23:23:23 +01:00
Chris Taylor
a580639abf
fix(agents): strip <thought> and <antthinking> tags from output
2026-01-10 23:19:58 +01:00
Peter Steinberger
494743a4e5
feat: run doctor after restart
2026-01-10 23:14:55 +01:00
Jonathan Wilkins
0afa370869
fix: add 'usage limit' to rate limit detection patterns
...
OpenAI/ChatGPT returns "You have hit your ChatGPT usage limit (plus plan)"
when users exceed their plan quota. This error wasn't being recognized as a
rate limit, so fallback to alternative models wasn't triggering.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com >
2026-01-10 23:11:06 +01:00
Peter Steinberger
08cc8f2281
refactor(agents): extract transcript repair module
2026-01-10 22:07:25 +00:00
Peter Steinberger
98377c7c6b
fix(agents): harden tool transcript repair
2026-01-10 21:45:15 +00:00
Peter Steinberger
805a29252e
test: add setup-token live smoke
2026-01-10 21:45:15 +00:00
Peter Steinberger
050c1c5391
fix(agents): include reasoningLevel in compaction params
2026-01-10 22:39:06 +01:00
Peter Steinberger
6d70524aa8
fix: add reasoning visibility hint
2026-01-10 22:24:22 +01:00
Peter Steinberger
aa30995aa1
test(live): add provider filters + google skip rules
2026-01-10 21:16:59 +00:00
Peter Steinberger
ab314a22e0
chore: refresh pi-ai patch repro note
2026-01-10 20:55:57 +00:00
Peter Steinberger
b9b1bc2726
test: relax reasoning replay expectations
2026-01-10 21:43:52 +01:00
Peter Steinberger
8a194b4abc
fix: align opencode-zen provider setup
2026-01-10 21:38:18 +01:00
Peter Steinberger
46e00ad5e7
fix: describe sandboxed elevated in prompt
2026-01-10 21:37:15 +01:00
Peter Steinberger
3389231ecb
feat(doctor): offer update first
2026-01-10 21:34:59 +01:00
Peter Steinberger
d772ff06c8
test: update openai responses reasoning replay
2026-01-10 21:20:26 +01:00
Peter Steinberger
cf192f8551
style: biome format
2026-01-10 19:47:17 +00:00
Peter Steinberger
d44bb41d27
fix: replay OpenAI reasoning for tool calls
2026-01-10 19:46:13 +00:00
Peter Steinberger
fa346d7b78
fix: accept Z_AI_API_KEY for zai
2026-01-10 19:46:13 +00:00
Peter Steinberger
9f9098406c
feat(sandbox): add sandbox explain inspector
2026-01-10 20:28:43 +01:00
Peter Steinberger
212b13b099
fix: repair tool-use history for anthropic
2026-01-10 19:15:57 +00:00
cristip73
c4e76eb635
fix: enable image attachments in chat messages for Claude API
...
Images were previously converted to markdown data URLs which Claude API
treats as plain text, not as actual images.
Changes:
- Add parseMessageWithAttachments() that returns {message, images[]}
- Pass images through the stack to session.prompt() as content blocks
- Filter null/empty attachments before parsing
- Strip data URL prefix if client sends it
This enables iOS and other clients to send images that Claude can actually see.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com >
2026-01-10 20:00:25 +01:00
Peter Steinberger
8c1d39064d
test: adjust tool id sanitization
2026-01-10 18:15:15 +00:00
Peter Steinberger
7c925aa5a0
style: format helper
2026-01-10 18:15:15 +00:00
Peter Steinberger
651a9e9be4
fix: restore minimax tool calling
2026-01-10 18:15:15 +00:00
Peter Steinberger
2d105d16f8
fix(opencode-zen): keep legacy aliases + rationale ( #669 ) (thanks @magimetal)
2026-01-10 18:25:43 +01:00
Magi Metal
738269eb74
feat(opencode-zen): update models with sensible defaults
...
- Replace model catalog with 11 current models: gpt-5.1-codex, claude-opus-4-5,
gemini-3-pro, alpha-glm-4.7, gpt-5.1-codex-mini, gpt-5.1, glm-4.7-free,
gemini-3-flash, gpt-5.1-codex-max, minimax-m2.1-free, gpt-5.2
- Add accurate per-token costs from OpenCode Zen pricing
- Add accurate context windows and output limits
- Update aliases for new model families (codex, glm, minimax)
- Remove deprecated models (sonnet, haiku, o-series, gemini-2.5)
2026-01-10 18:22:26 +01:00
