Agent DailyAgent Daily
releaseintermediate

[Release] openclaw/openclaw v2026.4.29-beta.4: openclaw 2026.4.29-beta.4

By steipetegithub
View original on github

OpenClaw v2026.4.29-beta.4 introduces significant enhancements across messaging, memory, provider coverage, and reliability. Key additions include active-run steering by default, a people-aware wiki with provenance views, NVIDIA provider onboarding, and improved gateway/channel reliability. Security improvements include OpenGrep scanning and sharper GHSA triage policies, while multiple channel fixes address Slack, Telegram, Discord, WhatsApp, and Teams edge cases.

Key Points

  • Messaging now defaults to active-run steering with 500ms followup debounce and supports visible-reply enforcement via global `messages.visibleReplies` config
  • Memory system evolves into people-aware wiki with provenance views, per-conversation Active Memory filters, and partial recall on timeout with bounded REM preview diagnostics
  • NVIDIA provider added with API-key onboarding, static catalog metadata, and literal model-ref picker support for hosted models
  • Bedrock Opus 4.7 thinking parity and safer Codex/OpenAI-compatible replay/streaming behavior implemented
  • Opt-in inferred follow-up commitments with heartbeat delivery, per-agent/per-channel scoping, and CLI management via `commitments.enabled`/`commitments.maxPerDay` config
  • Gateway reliability improved with slow-host startup handling, reusable model catalogs, event-loop readiness diagnostics, and stale-session recovery
  • Channel fixes address Slack Block Kit limits, Telegram proxy/webhook/polling resilience, Discord rate-limiting, WhatsApp delivery, and Teams/Matrix/Feistel edge cases
  • Security enhancements include OpenGrep scanning with precise rulepack, GHSA triage policy refinement, and safer exec/pairing/owner-scope handling
  • SQLite-backed plugin state store added for restart-safe keyed registries with TTL and automatic plugin isolation
  • Tool sections (`tools.exec`, `tools.fs`) no longer implicitly widen restrictive profiles; explicit `alsoAllow` entries required with startup warnings for affected configs

Found this useful? Add it to a playbook for a step-by-step implementation guide.

Workflow Diagram

Start Process
Step A
Step B
Step C
Complete
Quality

Concepts

Artifacts (5)

commitments_configyamlconfig
commitments:
  enabled: true
  maxPerDay: <number>
  heartbeatInterval: <duration>
messages_configyamlconfig
messages:
  visibleReplies: true
  queue: steer
  followupDebounce: 500ms
  groupChat:
    visibleReplies: true
active_memory_configyamlconfig
activeMemory:
  allowedChatIds:
    - <chat_id_1>
    - <chat_id_2>
  deniedChatIds:
    - <chat_id_3>
  partialRecallTimeout: <duration>
docker_onboarding_skipbashcommand
docker run -e OPENCLAW_SKIP_ONBOARDING=true openclaw:v2026.4.29-beta.4
tool_restrictions_configyamlconfig
tools:
  exec:
    alsoAllow:
      - <permission_1>
  fs:
    alsoAllow:
      - <permission_2>
profile: minimal
[Release] openclaw/openclaw v2026.4.29-beta.4: openclaw 2026.4.29-beta.4 | Agent Daily