acp-router

acp-router is an Other agent platforms skill for AI agents, published by openclaw in the openclaw repository on GitHub. Skillspace rates acp-router D (65/100) and the safety scan flagged it for review; a judge model has also reviewed it, and it has been tested head-to-head on real tasks.

Route plain-language requests for Claude Code, Cursor, Copilot, OpenClaw ACP, OpenCode, Gemini CLI, Qwen, Kiro, Kimi, iFlow, Factory Droid, Kilocode, or explicit ACP harness work into either OpenClaw ACP runtime sessions or direct acpx-driven sessions ("telephone game" flow). For coding-agent thread requests, read this skill first, then use only `sessions_spawn` for thread creation. Codex chat binding defaults to the native Codex app-server plugin unless ACP is explicit or background spawn needs ACP.

Grade
D
Quality
65 / 100
Safety
Needs review
GitHub stars
383,040

Won 4 of 6 blind task comparisons against the category's other leading skill, a deliberately weak skill and a no-skill baseline. Placed #5 of 6 in the judged category final for Other agent platforms.

Our two independent classifiers differ on this skill's category, so treat the placement as provisional.

This rating describes the version of the repository we crawled; it may have changed since, so go by its current files.

Our review

Architecture
95 / 100
Maturity
Basic: a single SKILL.md
Security scan
Medium-risk patterns
License
Not declared. Check the repository before reusing it.
Adoption
Adopted by 0 other repositories, in 0 collections
Scope
Vendor-specific: built around one product. Routing via OpenClaw ACP product runtime and acpx

Head-to-head

Won 4 of 6 blind task comparisons against the category's other leading skill, a deliberately weak skill and a no-skill baseline.

Judge review

Placed #5 of 6 in the judged category final for Other agent platforms.

  • Trigger8080
  • Instructions9090
  • Specificity9292
  • Safety7272
  • Disclosure5555
  • Maintenance5555
Highly detailed multi-harness router with real recovery and retry procedures and pinned-version resolution, but it is a dense single file that authorizes self-repair actions with only partial confirmation gating.

What the scanner flagged

  • reads environment variables

Not checked yet

  • Sibling files (scripts, references) not fetched; security scan covers SKILL.md only

Skillspace indexes and rates this skill; it does not host or endorse it. Run third-party skills at your own risk.