agent/pi - pi - Penguin

Release v0.55.4

Mario Zechner · 2026-03-02 23:24:27 +01:00

v0.55.4 9a53d950c4

fix(coding-agent): close retry wait race across queued events (from #1726 )

Mario Zechner · 2026-03-02 23:20:16 +01:00

8903299074

docs(coding-agent): add unreleased new features section

Mario Zechner · 2026-03-02 23:08:08 +01:00

2e7ec46176

Update README.md

Mario Zechner · 2026-03-02 23:00:29 +01:00

5ec0540bf4

OSS vacation off

Mario Zechner · 2026-03-02 23:00:15 +01:00

a2b52f5cee

chore(coding-agent): add changelog entry for PR #1719\n\ncloses #1719

Mario Zechner · 2026-03-02 22:59:32 +01:00

8e06afbf34

merge: PR #1719 for local testing

Mario Zechner · 2026-03-02 22:54:58 +01:00

7b7b967aef

docs(coding-agent): fix changelog attribution format

Mario Zechner · 2026-03-02 22:52:55 +01:00

8032a302e9

docs(coding-agent): reference #1237 in changelog attribution

Mario Zechner · 2026-03-02 22:51:36 +01:00

e18c74dda0

docs(coding-agent): add changelog attribution for #1720

Mario Zechner · 2026-03-02 22:51:00 +01:00

8344f3da49

fix(coding-agent): add tool promptGuidelines support fixes #1720

Mario Zechner · 2026-03-02 22:50:08 +01:00

8d4a49487a

fix(coding-agent): support dynamic tool registration and tool prompt snippets closes #1720

Mario Zechner · 2026-03-02 22:32:07 +01:00

bc2fa8d6d0

Restore approved contributors

Mario Zechner · 2026-03-02 20:50:49 +01:00

ca5510158d

fix(coding-agent): serialize session event handling to preserve message order (fixes #1717 )

Mario Zechner · 2026-03-02 20:48:22 +01:00

dfc779faab

docs(mom): add changelog entry for settings manager fix

closes #1444

Mario Zechner · 2026-03-02 19:43:35 +01:00

062f7ff52d

chore(ai): update generated models data

Mario Zechner · 2026-03-02 19:41:38 +01:00

105e296b69

refactor(mom): use SettingsManager with workspace storage

fixes #1444

Mario Zechner · 2026-03-02 19:41:38 +01:00

46bfa7931e

chore: approve contributor alasano

github-actions[bot] · 2026-03-02 18:27:03 +00:00

0c23a3415e

fix(coding-agent): remove extra spacer before streaming tool blocks

Aljosa Asanovic · 2026-03-02 10:22:51 -05:00

b97310474b

fix(coding-agent): allow suppressing custom tool transcript blocks

Aljosa Asanovic · 2026-03-02 10:22:41 -05:00

480d6bc62d

Add [Unreleased] section for next cycle

Mario Zechner · 2026-02-27 22:54:44 +01:00

95276df060

Release v0.55.3

Mario Zechner · 2026-02-27 22:54:03 +01:00

v0.55.3 9d68043cbc

chore(coding-agent): add changelog entry for windows image paste keybinding closes #1682

Mario Zechner · 2026-02-27 22:51:51 +01:00

af85b715db

fix(coding-agent): use alt+v for image pasting on Windows (#1682 )

Duncan Ogilvie · 2026-02-27 22:51:16 +01:00

cec8c7888b

Add [Unreleased] section for next cycle

Mario Zechner · 2026-02-27 22:26:46 +01:00

edca5dd769

Release v0.55.2

Mario Zechner · 2026-02-27 22:26:00 +01:00

v0.55.2 10180037d2

docs: audit and add missing changelog entries for v0.55.1..HEAD

Mario Zechner · 2026-02-27 22:25:03 +01:00

15e0a74a3c

fix(coding-agent): prevent duplicate session headers when forking from pre-assistant entry

createBranchedSession() wrote the file and set flushed=true even when the
branched path had no assistant message. The next _persist() call saw no
assistant, reset flushed=false, and the subsequent flush appended all
in-memory entries to the already-populated file, duplicating the header
and entries.

Fix: defer file creation when the branched path has no assistant message,
matching the newSession() contract. _persist() creates the file on the
first assistant response.

closes #1672

Mario Zechner · 2026-02-27 22:18:26 +01:00

2f64df1e52

fix(ai): handle redacted_thinking blocks, skip interleaved beta for adaptive models, drop temperature with thinking

- Map redacted_thinking to ThinkingContent with redacted: true instead of
  adding a new content type. The opaque payload goes in thinkingSignature,
  thinking text is set to "[Reasoning redacted]" so it renders naturally
  everywhere. Cross-model transform drops redacted blocks.
- Skip interleaved-thinking-2025-05-14 beta header for Opus 4.6 / Sonnet 4.6
  where adaptive thinking makes it deprecated/redundant.
- Do not send temperature when thinkingEnabled is true (incompatible with
  both adaptive and budget-based thinking).

Based on #1665 by @tctev

Mario Zechner · 2026-02-27 21:53:25 +01:00

9825c13f5f

Closes #1671 , switch to GLM-5 for tests

Mario Zechner · 2026-02-27 21:35:40 +01:00

afe9ae06e8

docs(ai): add changelog entry for Z.ai thinking fix #1674

Mario Zechner · 2026-02-27 21:33:42 +01:00

187a99c720

fix(ai): use enable_thinking for Z.ai instead of thinking param (#1674 )

Z.ai uses the same enable_thinking: boolean parameter as Qwen to control reasoning, not thinking: { type: "enabled" | "disabled" }.

The wrong parameter name means Z.ai ignores the disable request and always runs with thinking enabled, wasting tokens and adding latency.

Merge the Z.ai and Qwen branches since they use the same format.

PR by @okuyam2y

Yoshiaki Okuyama · 2026-02-27 21:33:14 +01:00

22b3be834e

Manual merge of #1677

Mario Zechner · 2026-02-27 21:27:55 +01:00

8880cd3cb8

Merge branch 'pr-1669-fixes'

# Conflicts:
#	package-lock.json
#	packages/ai/CHANGELOG.md
#	packages/coding-agent/CHANGELOG.md

Mario Zechner · 2026-02-27 21:04:00 +01:00

3dcb3c1c77

fix(coding-agent,ai): finalize provider unregister lifecycle and dependency security updates fixes #1669

Mario Zechner · 2026-02-27 21:00:25 +01:00

2f55890452

fix(coding-agent): ignore SIGINT while process is suspended (#1668 )

Add a no-op SIGINT handler before sending SIGTSTP in handleCtrlZ().
Remove it on SIGCONT. Prevents Ctrl+C from killing the backgrounded
process.

Aliou Diallo · 2026-02-27 13:51:45 +01:00

0b4dea466d

Add [Unreleased] section for next cycle

Mario Zechner · 2026-02-26 01:10:55 +01:00

9a0a8d7ccb

Release v0.55.1

Mario Zechner · 2026-02-26 01:10:09 +01:00

v0.55.1 a753e24135

docs(changelog): cover upstream unreleased ai and coding-agent entries

Mario Zechner · 2026-02-26 01:13:32 +01:00

a27ea5b9e2

docs(changelog): audit unreleased entries and add missing items

Mario Zechner · 2026-02-26 01:11:34 +01:00

b2983c483e

chore: add repository .gitattributes for line ending normalization

Mario Zechner · 2026-02-26 01:11:34 +01:00

43c3c99dde

fix syntax for initializing AuthStorage in docs (#1539 )

Nader Dabit · 2026-02-26 00:35:13 +01:00

40a8706bb9

fix(ai): enable adaptive thinking for sonnet 4.6 and clamp xhigh effort (#1548 )

* fix(ai): enable adaptive thinking for sonnet 4.6 and clamp xhigh effort

* chore(ai): drop changelog edit from contribution

---------

Co-authored-by: tctev <224793535+tctev@users.noreply.github.com>

tctev · 2026-02-26 00:34:06 +01:00

e9d0074fa6

fix(ai): don't cache false for Vertex ADC credentials during async import race (#1550 )

`hasVertexAdcCredentials()` uses dynamic imports to load `node:fs`,
`node:os`, and `node:path` to avoid breaking browser/Vite builds. These
imports are fired eagerly but resolve asynchronously. If the function is
called during gateway startup before those promises resolve, `_existsSync`,
`_homedir`, and `_join` are still null — causing the function to cache
`false` permanently and never re-evaluate.

This means users with valid `GOOGLE_APPLICATION_CREDENTIALS`,
`GOOGLE_CLOUD_PROJECT`, and `GOOGLE_CLOUD_LOCATION` configured are silently
treated as unauthenticated for Vertex AI. Calls fall back to the AI Studio
endpoint (generativelanguage.googleapis.com) which has much stricter rate
limits, causing unexpected 429 errors even though Vertex credentials are
correctly configured.

Fix: in Node.js/Bun environments, return false without caching when the
async modules aren't loaded yet, so the next call retries. Only cache false
permanently in browser environments where `fs` is genuinely unavailable.

Co-authored-by: Jeremiah Gaylord <jeremiahgaylord-web@users.noreply.github.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

jeremiahgaylord-web · 2026-02-26 00:31:16 +01:00

cf656c169c

fix(subagent): use getAgentDir for user agents path (#1559 )

Tianshu Wang · 2026-02-26 00:22:42 +01:00

7390f830de

feat(ai): add gemini-3.1-pro-preview to google-gemini-cli provider (#1599 )

Add Gemini 3.1 Pro Preview model to the Cloud Code Assist (google-gemini-cli)
provider for parity with the google and google-vertex providers that already
include this model.

Tested and confirmed working via the Cloud Code Assist API endpoint.