MCP and Anthropic recovery pathways have also been hardened.
Peter Steinberger 🦞#495
🦞 Big security update for users and enterprise: Custom install policy for plugins and skills! 🔒
OpenClaw will call an executable of your choice with an extensive profile and source code of every plugin/skill.
You make the decision to allow install.
Many users praised OpenClaw's new parallel web search, keyless access, and security policies for improving reliability and agent infrastructure, while a few criticized the versioning changes as random.
security.installPolicy lets operators approve or block skill/plugin installs with a trusted local command. It runs after staging and fails closed when enabled but unavailable. https://docs.openclaw.ai/tools/skills-config#operator-install-policy-securityinstallpolicy
Parallel is now bundled as a web_search provider. Use the zero-config free path or bring PARALLEL_API_KEY for the paid API, with dense excerpts built for LLM context. https://docs.openclaw.ai/tools/parallel-search
Release naming changed: OpenClaw now uses YYYY.M.PATCH. That last number is a monthly patch counter, not the calendar day; old tags still compare correctly. This is to help support future critical fixes on prior releases. https://docs.openclaw.ai/reference/RELEASING
Default for new, <10 seconds to switch your existing 🦞 - free and keyless
The best web search for AI is now the keyless default in @openclaw
Free, no account, LLM-optimized. Just powerful web search for your claw 🦞
Anthropic extended-thinking sessions recover better after prompt-cache expiry or Gateway restart, and richer MCP tool blocks get normalized before they poison history. https://docs.openclaw.ai/providers/anthropic
QQBot now strips model reasoning/thinking scaffolding before native delivery, so users see the answer instead of raw internal narration. https://docs.openclaw.ai/channels/qqbot
@openclaw
@openclaw OpenClaw on it's way to the next update already:
Plugin and ClawHub installs are sturdier: pinned GitHub skill commits, install-policy checks, and SQLite-backed install records for better update/repair behavior. https://docs.openclaw.ai/tools/plugin
Matrix got less awkward: voice notes can be transcribed before mention gating, and native threads preserve reply/read context instead of flattening the conversation. https://docs.openclaw.ai/channels/matrix
@openclaw 🌐🤝🦞
@openclaw @grok can openclaw be used with a claude account login or just api usage?
@openclaw https://hvtracker.net/agents/openclaw/
@openclaw next: memory :)
@openclaw Can’t wait to use the new web search
@openclaw Lol no fable?
@openclaw Parallel search plus safer install policy is a nice combo, because agents get useful faster without turning plugins into a trust fall.
@openclaw Update coming with Fable soon?
@grok @openclaw @grok so I can use it without paying extra usage from my max account? Or only available with extra usage?
The pre-June 15 policy is Anthropic's current rule (until June 15 2026). CLI login in OpenClaw reuses your full Max subscription quotas and rate limits for programmatic/agent use exactly like normal chat—no separate charges yet.
After June 15 they split the pools. CLI / claude -p / third-party Agent SDK usage switches to a dedicated monthly credit first ($100 for Max 5x, $200 for Max 20x). Once that's gone it either stops or bills at full API rates if you enable extra usage. Interactive chat stays on your regular sub limits.
That's why it currently works inside your existing plan. Check openclaw docs or Anthropic support for the exact credit amounts on your tier.
MCP and Anthropic recovery pathways have also been hardened.
Peter Steinberger 🦞#495🦞 Big security update for users and enterprise: Custom install policy for plugins and skills! 🔒
OpenClaw will call an executable of your choice with an extensive profile and source code of every plugin/skill.
You make the decision to allow install.
