Harness detection
SkillAI & modelsLets your agent detect which local coding harnesses are installed, including their versions and account status.
Available today. Use it from your connected AI after setup.
No other account needed.
Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.
Then ask your AI: use the Harness detection skill
About this skill
Change local coding-harness discovery, account status, model flag detection, or version probes.
What this skill tells your AI
The instructions your AI receives, as published by smithersai/smithers in .agents/skills/smithers-harness-detect/SKILL.md and read by ahel’s review.
Read the harness-detect README and the live app adapter apps/app/src/bun/Harnesses.ts before changing a detector. The package is pure over an injected HarnessHost; filesystem and process probes belong in the adapter. Keep binary search, account status, and model flags grounded in the installed harness's actual help and observed files.
A version probe runs only --version with probeEnv's PROBE_ENV_KEYS allowlist. Do not pass session or vendor credentials to the child; the live adapter also owns sandboxing and timeout. The Codex JWT email claim is decoded without verification and must not be returned as an authenticated identity. Test the injected host with fixture paths and a probe environment that excludes unrelated tokens.
Signals
- GitHub stars
- 422
- Forks
- 50
- Last commit
- Sep 2026
Advanced
- Catalog kind
- skill
- Key
smithers-harness-detect- Source
- github.com/smithersai/smithers