Commit graph

663 commits

Author SHA1 Message Date
Melih Mucuk
cceb5908d9
fix: opencode provider uses system role instead of developer (#755)
* fix: opencode provider uses system role instead of developer for /v1 endpoint

* changelog updated
2026-01-15 21:26:31 +01:00
Mario Zechner
e46cc55fe0 Add [Unreleased] section for next cycle 2026-01-15 17:36:40 +01:00
Mario Zechner
0aaa4ef5e4 Release v0.46.0 2026-01-15 17:35:54 +01:00
Mario Zechner
a216cc496f chore: update changelogs 2026-01-15 17:34:59 +01:00
Roshan Singh
b18f401d9e
fix(ai): avoid unsigned Gemini 3 tool calls (#741) 2026-01-15 13:12:39 +01:00
Aadish Verma
cd83e73845
feat: add gpt-5.2-codex models for copilot & zen (#734) 2026-01-15 03:24:20 +01:00
Mario Zechner
cc8c51d9ae chore(ai): regenerate models 2026-01-14 22:25:21 +01:00
Mario Zechner
0a7537bf86 Revert "feat(ai): add gpt-5.2-codex to OpenAI provider (#730)"
This reverts commit 5a795b9857.
2026-01-14 22:22:55 +01:00
Anton
5a795b9857
feat(ai): add gpt-5.2-codex to OpenAI provider (#730)
* feat(ai): add gpt-5.2-codex to OpenAI provider

* fix(ai): avoid build break when model generation misses providers
2026-01-14 22:21:01 +01:00
Burak Varlı
9a438465eb
fix(ai): signature support for non-Anthropic models in Amazon Bedrock provider (#727)
* Add Amazon Bedrock models test suite for agent package

Tests basic prompts, multi-turn conversations with thinking, and
synthetic thinking signatures across all Bedrock models.

Known issues are categorized and skipped:
- Models requiring inference profile (5)
- Invalid model IDs for us-east-1 region (6)
- Max tokens config exceeds model limit (2)
- No signature support in reasoningContent (10)
- Rejects reasoning content in user messages (25)
- Validates signature format - Anthropic newer models (7)

* Fix Bedrock signature support for non-Anthropic models

Only include the signature field in reasoningContent.reasoningText for
Anthropic Claude models. Other models (OpenAI, Qwen, Minimax, Moonshot,
etc.) reject this field with:
"This model doesn't support the reasoningContent.reasoningText.signature field"

This fix enables multi-turn conversations with thinking content for
10 additional Bedrock models that previously failed.

https://buildwithpi.ai/session?7e39c05f66ea358da3f993c267fe3e29

* Add a CHANGELOG entry
2026-01-14 19:21:35 +01:00
Jian Zhang
558a77b45f
feat(ai): add support for MiniMax China (minimax-cn) provider (#725)
Co-authored-by: Jian Zhang <jzhang@yanhuangdata.com>
2026-01-14 15:41:47 +01:00
Mario Zechner
95859725b7 Add [Unreleased] section for next cycle 2026-01-14 00:11:36 +01:00
Mario Zechner
d08821cc4a Release v0.45.7 2026-01-14 00:10:52 +01:00
Mario Zechner
c16698130c chore: update changelogs 2026-01-14 00:09:51 +01:00
Markus Ylisiurunen
653025e6ca
Fix OpenAI responses timeout option (#706) 2026-01-14 00:07:11 +01:00
Pablo Tovar
b74535dc85
fix: apply message transforms for bedrock tool calls (#707)
Ensure Bedrock uses transformMessages before conversion.
2026-01-14 00:05:47 +01:00
Mario Zechner
d83e1745d8 Add [Unreleased] section for next cycle 2026-01-13 22:48:47 +01:00
Mario Zechner
f8dc72d24f Release v0.45.6 2026-01-13 22:48:05 +01:00
Mario Zechner
3fe73ecd04 docs(ai): add parseStreamingJson export fix to changelog 2026-01-13 22:42:45 +01:00
Mario Zechner
2b75948c2e fix(ai): export parseStreamingJson from main package for tsx compatibility 2026-01-13 22:08:19 +01:00
Mario Zechner
7d45e434de Merge branch 'main' into feat/tui-overlay-options 2026-01-13 22:06:02 +01:00
Mario Zechner
9994ebbedd Add [Unreleased] section for next cycle 2026-01-13 21:47:38 +01:00
Mario Zechner
9a08bb0d1e Release v0.45.5 2026-01-13 21:46:57 +01:00
Mario Zechner
3a1228c0b5 Add [Unreleased] section for next cycle 2026-01-13 21:37:34 +01:00
Mario Zechner
bb0ba88b97 Release v0.45.4 2026-01-13 21:36:52 +01:00
Mario Zechner
09d409cc92 Fix z.ai thinking/reasoning params, fixes #688
Z.ai uses thinking: { type: "enabled" | "disabled" } instead of
OpenAI's reasoning_effort. Added thinkingFormat compat flag to handle
this. Thinking is now explicitly enabled/disabled based on user setting.
2026-01-13 18:34:07 +01:00
Markus Ylisiurunen
00ba005e50
set the prompt cache key to session id (#698) 2026-01-13 18:29:36 +01:00
Mario Zechner
28072cb31f Add more models to stream.test.ts for Vercel, set infinite timeout on OpenAI responses, closes #690 2026-01-13 17:08:56 +01:00
Mario Zechner
1d8d5d3d88 Add handleImage and multiTurn tests for vercel-ai-gateway in stream.test.ts 2026-01-13 16:56:48 +01:00
Mario Zechner
19f3c23f6d Fix PR #689: Add changelog attribution, coding-agent changelog, fix test types, add provider to test suites
- Fix ai/CHANGELOG.md: add PR link and author attribution
- Add coding-agent/CHANGELOG.md entry for vercel-ai-gateway provider
- Fix model-resolver.test.ts: use anthropic-messages API type to match generated models
- Add vercel-ai-gateway to test suites: tokens, abort, empty, context-overflow, unicode-surrogate, tool-call-without-result, image-tool-result, total-tokens, image-limits
2026-01-13 16:46:00 +01:00
Timo Lins
dd263f3c24 Update model list 2026-01-13 16:42:34 +01:00
Timo Lins
65eb738c90 Rename to vercel-ai-gateway for clarity 2026-01-13 16:42:34 +01:00
Timo Lins
9860ee86f3 Change to Anthropic compatible API
It seemed as if the OpenAI message spec tried to send non-compliant messages with { text: "" } instead of { contet: "" }, which the AI Gateway did not accept.
2026-01-13 16:42:34 +01:00
Timo Lins
164a69a601 Add Vercel AI Gateway support 2026-01-13 16:42:34 +01:00
Markus Ylisiurunen
922b0a4668
add eu cross-region inference model ids for anthropic models (#685) 2026-01-13 13:02:27 +01:00
Nico Bailon
a4ccff382c feat(tui): overlay positioning API with CSS-like values
Add OverlayOptions for configurable positioning (anchor, margins, offsets,
percentages). Add OverlayHandle for programmatic visibility control with
hide/setHidden/isHidden. Add visible callback for responsive overlays.

Extension API: ctx.ui.custom() now accepts overlayOptions and onHandle callback.

Examples: overlay-qa-tests.ts (10 test commands), doom-overlay (DOOM at 35 FPS).
2026-01-12 22:44:58 -08:00
Mario Zechner
75146a2b96 Add [Unreleased] section for next cycle 2026-01-13 04:57:25 +01:00
Mario Zechner
2090599e4b Release v0.45.3 2026-01-13 04:56:41 +01:00
Mario Zechner
3c60ffa677 Fix tool call ID normalization for cross-provider switches to Anthropic/GitHub Copilot 2026-01-13 04:07:10 +01:00
Mario Zechner
6c2167a0fe Add [Unreleased] section for next cycle 2026-01-13 03:28:40 +01:00
Mario Zechner
c55082bb66 Release v0.45.2 2026-01-13 03:27:56 +01:00
Mario Zechner
71ee79f31a Add [Unreleased] section for next cycle 2026-01-13 03:07:28 +01:00
Mario Zechner
e4a98d8381 Release v0.45.1 2026-01-13 03:06:43 +01:00
Mario Zechner
3efe58a62c Add [Unreleased] section for next cycle 2026-01-13 02:42:20 +01:00
Mario Zechner
e22feba494 Release v0.45.0 2026-01-13 02:41:36 +01:00
Mario Zechner
8af8d0d672 Add MiniMax provider support (#656 by @dannote)
- Add minimax to KnownProvider and Api types
- Add MINIMAX_API_KEY to getEnvApiKey()
- Generate MiniMax-M2 and MiniMax-M2.1 models
- Add context overflow detection pattern
- Add tests to all required test files
- Update README and CHANGELOG with attribution

Also fixes:
- Bedrock duplicate toolResult ID when content has multiple blocks
- Sandbox extension unused parameter lint warning
2026-01-13 02:27:09 +01:00
Mario Zechner
9c694e36f8 Add changelog entry for #670 2026-01-13 01:05:26 +01:00
Ahmed Kamal
ff15414258
Improve Gemini CLI provider retries and headers (#670)
Improve Gemini CLI provider retries and headers

- Add Antigravity endpoint fallback (tries daily sandbox then prod when baseUrl is unset)
- Parse retry delays from headers (Retry-After, x-ratelimit-reset, x-ratelimit-reset-after) before body parsing
- Derive stable sessionId from first user message for cache affinity
- Retry empty SSE streams with backoff without duplicate start/done events
- Add anthropic-beta header for Claude thinking models only
2026-01-13 01:04:53 +01:00
Danila Poyarkov
9e4ae98358
Improve Google Cloud Code Assist error handling (#665)
* Improve Cloud Code Assist error messages

- Extract just the message from verbose JSON error responses
- Extract cause from generic 'fetch failed' errors for better diagnostics

* Make 'other side closed' network error retryable

* Make 'other side closed' network error retryable
2026-01-13 00:41:20 +01:00
Mario Zechner
259a41c8fd docs: Note prompt caching in Bedrock changelog entry 2026-01-13 00:38:35 +01:00