RepoJournal
AI Agents AI Agents
47 wires and counting

$ follow AI Agents

Keep up with AI Agents in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-09-05
stories 23

© 2026 RepoJournal Home Showcase Explore How it works Privacy

$ the-wire · showcase

LangChain adds async OpenAI tools; OpenHands fixes flaky CI

By RepoJournal · Filed · About AI Agents

LangChain core 1.6.2 and new OpenAI async tool support land alongside OpenHands fixes for a flaky test suite and broken branch URLs.

LangChain released langchain-core 1.6.2 [1], which includes a new feature: async tool support for OpenAI [2]. This lets developers call tools asynchronously, improving performance in concurrent or event-driven applications. The release also bumps mistune to 3.3.3 and tornado to 6.5.8 in libs/core [1]. Additionally, OpenAI integration gains support for `configuration_update` [3], allowing reasoning effort changes mid-conversation, as documented by OpenAI.

OpenHands fixed an intermittent CI failure where tests passed but exit code was 1 due to an unhandled rejection of ProgressEvent after jsdom teardown [4]. The fix keeps ProgressEvent resolvable, stabilizing the test-and-build pipeline. Separately, OpenHands now percent-encodes branch names in provider branch URLs [5], fixing broken links for branches containing special characters like '#' (e.g., feature/ui#123). The UI also displays message timestamps on hover or keyboard focus [6].

Cline fixed a bug where apply_patch's Add File operation silently overwrote existing files [7]. The guard intended to throw 'Add File Error: File already exists' was dead code because ADD targets were never loaded into the map. Now loadFiles includes ADD targets, so the guard fires correctly. This prevents accidental data loss during patching. Also, the model picker search now keeps section headers visible [8], clarifying which tier a model belongs to when duplicate names appear.

Action items

References

  1. [1] langchain-core==1.6.2 ↗ langchain-ai/langchain
  2. [2] feat(openai): support async tools ↗ langchain-ai/langchain
  3. [3] feat(openai): support `configuration_update` ↗ langchain-ai/langchain
  4. [4] fix(test): keep ProgressEvent resolvable after jsdom teardown ↗ OpenHands/OpenHands
  5. [5] fix(utils): percent-encode branch names in provider branch URLs ↗ OpenHands/OpenHands
  6. [6] feat: display message timestamps on hover ↗ OpenHands/OpenHands
  7. [7] fix(core): make apply_patch Add File refuse to overwrite existing files ↗ cline/cline
  8. [8] ui: keep section headers visible while searching in model picker ↗ cline/cline

Quick answers

What shipped in AI Agents on September 5, 2026?
LangChain core 1.6.2 and new OpenAI async tool support land alongside OpenHands fixes for a flaky test suite and broken branch URLs. In total, 11 commits, 11 pull requests, and 1 releases landed.
Who contributed to AI Agents on September 5, 2026?
9 developers shipped this update, including github-actions[bot], ccurme, npentrel, langchain-oss-model-profiles[bot], lzhan011, LosAveRoad, marmar9615-cloud, and imitater-dou, and 1 more.
What were the notable AI Agents updates?
langchain-core==1.6.2, feat(openai): support async tools, and feat(openai): support `configuration_update`.