44 wires and counting

$ follow OpenClaw

Keep up with OpenClaw 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 2026-06-08
stories 60

© 2026 RepoJournal Home Showcase Explore How it works Privacy

$ the-wire · showcase

Leaner tool catalogs, cleaner Anthropic replays

By RepoJournal · Filed · About OpenClaw

openclaw/openclaw compacted default local tool catalogs and dropped persistent reasoning signatures from Anthropic replays.

Default localModelLean now compacts the Tool Search controls when the operator has not configured tools.toolSearch, while explicit Tool Search settings and direct message-tool delivery semantics are preserved [1]. If you rely on the default lean model without a custom tool search config, expect the UI to present a more compact catalog; explicit settings remain untouched.

Anthropic replay payloads no longer carry persisted OpenAI-compatible reasoning_content thinking placeholders. The fix filters those signatures from direct replay payloads and updates the focused Anthropic provider test [2]. Expect outbound Anthropic requests to be cleaner, without the thinkingSignature noise that previously surfaced.

A new shared live provider catalog runtime now routes live model discovery for OpenAI, xAI, OpenCode Go, Chutes, DeepInfra, Venice, NVIDIA, and Vercel AI Gateway through one helper [3]. Duplicated provider-local live catalog caching is removed, and auth marker stripping, empty live-result retries, and OpenAI custom-base-url handling are hardened. Also, ReplyChain and ReplyToBody JSON paths now apply head+tail body truncation instead of prefix-only truncation, using code-point iteration to avoid splitting UTF-16 surrogate pairs [4]. On iOS, when gateway realtime mode fails, the app now falls back to native speech recognition/TTS with a single generic gateway/runtime issue and a fallback banner showing configured versus active talk mode [5].

Action items

References

  1. [1] fix(agents): compact lean local tool catalogs ↗ openclaw/openclaw
  2. [2] fix(anthropic): drop reasoning_content replay signatures (#91231) ↗ openclaw/openclaw
  3. [3] feat: add live provider model catalog helper ↗ openclaw/openclaw
  4. [4] fix(inbound-meta): apply head+tail body truncation to ReplyChain and ReplyToBody JSON paths (fixes #91042) ↗ openclaw/openclaw
  5. [5] feat(ios): clarify talk realtime fallback ↗ openclaw/openclaw

Quick answers

What shipped in OpenClaw on June 8, 2026?
openclaw/openclaw compacted default local tool catalogs and dropped persistent reasoning signatures from Anthropic replays. In total, 38 commits and 22 pull requests landed.
Who contributed to OpenClaw on June 8, 2026?
6 developers shipped this update, including Vincent Koc, Yzx, Jason (Json), zenglingbiao, ngutman, and openclaw-docs-sync[bot].
What were the notable OpenClaw updates?
fix(agents): compact lean local tool catalogs, fix(anthropic): drop reasoning_content replay signatures (#91231), and feat: add live provider model catalog helper.