v0.22.0
openclaw/wacliv0.22.0Oct 16, 2026by github-actions[bot]
Upgrading?
AI Summary
This release introduces bounded search contexts and cursor reads for messages, along with a content-free incremental journal. It enhances chat recovery, safer send mechanisms, and includes five networking security fixes. The update also refines media handling, sync logic, and dependency management.
Key Highlights
- Bounded message search context and cursor reads
- Content-free incremental journal
- Safer sends and chat recovery
- Five networking security fixes
- Updates to Go and Node.js dependencies
New Features
- Bounded chronological search context with exact local chat/contact name filters
- Bounded cursor reads for list, search, and export with archive invalidation
- Content-free incremental journal, bounded drains, and retention expiry
- Exact inbound mark-read boundary preserving newer unread work
- Retention of native quote IDs and canonical sender identities in history
- Reuse of verified cached downloads
- Exposure of nonpublishing dry run for dependency and toolchain verification
Full Release Notes
## 0.22.0 - 2026-10-16 **Highlights:** bounded message search context and cursor reads, a content-free incremental journal, safer sends and chat recovery, and five networking security fixes. - Messages: add bounded chronological search context and exact local chat/contact name filters with explicit ambiguity errors. - Messages: add bounded cursor reads for list, search, and export, with explicit archive invalidation and shared successful-sync freshness. - Messages: add a content-free incremental journal, bounded drains, retention expiry, and identity reconciliation. - Dependencies: update Go networking, crypto, terminal, protocol and parser modules, pnpm 12.13.0, and the dead-code checker after the 48-hour cooldown; fix five reachable x/net advisories while retaining Go and Node minimums. - Chats: add an exact inbound mark-read boundary that preserves newer unread work through receipts, sync echoes, and recovery replay. Thanks @artemgetmann (#488). - Chats: let `mark-read --receipts` clear local unread metadata when no eligible incoming message is stored, report the zero-receipt result as local-only, and preserve arrivals during the clear. Thanks @nicolasbiondic (#491). - Media: reuse verified cached downloads and add bounded exact phone recovery through the running sync owner. Thanks @artemgetmann (#482). - Send: avoid fresh message retries after ambiguous disconnects while retaining safe lookup and presence retries. Thanks @artemgetmann (#483). - Send: retain native quote IDs and canonical sender identities in immediate text and attachment history, and save attachments under the canonical recipient. Thanks @artemgetmann (#485, #486). - Messages: preserve exact raw text, captions, and edits, including surrounding whitespace. Thanks @artemgetmann (#484). - Sync: count the first decoded group payload after a protocol placeholder once and preserve it across empty protocol replays without discarding explicit empty edits. Thanks @artemgetmann (#487). - Chats: finish an app-state recovery the primary answers after the waiter timed out, persisting the late snapshot and clearing that attempt's replay markers. Thanks @nicolasbiondic (#489). - Chats: on collections whose full sync is known not to verify, let chat-state writes proceed after an incremental sync instead of a failing full replay and a five-minute wait for the phone; only the pre-write attempt marker is cleared, while older recovery intents and each write's post-send intent are kept until a verified full replay or a primary recovery snapshot covers them. Thanks @nicolasbiondic (#490). - Release: expose the shared workflow’s nonpublishing dry run for dependency and toolchain verification.
Related Repositories
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
A fast reverse proxy to help you expose a local server behind a NAT or firewall to the internet.
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for LLMs
The world’s fastest framework for building websites.
:cherry_blossom: A command-line fuzzy finder