agentleFS
0xrizz0xrizz/os-manager/.agents/skills/harness-check/SKILL.md

harness-check

Skill1 starsChanged 3 months agoMITView on GitHub

Use when verifying Claude Code harness integrity, running the security hook test suite, validating multi-agent skill symlinks, or checking settings.json syntax

  • Deletes or force-pushes

The file

Files
SKILL.md

Harness Check Skill

End-to-end self-diagnostic verification engine for the Claude Code agent harness, security hooks, multi-agent symlinks, and configuration validity.

Trigger Scenarios

  • Verifying harness integrity after modifying scripts, lifecycle hooks, or configuration
  • Validating deterministic security guardrails and pre/post tool execution filters
  • Checking multi-agent skill symlink synchronization across .claude/, .agents/, and ~/.gemini/
  • Periodic health checks of the developer automation matrix

Invocation

${PROJECT_DIR:-.}/scripts/harness_check.sh

Validation Matrix

The harness check runs three primary validation pillars:

  1. Hook & Guardrail Unit Test Suite (tests/test_harness.sh):
    • Deterministic Tier 3 invariant blocking (rm -rf /, rm -rf ~) with Exit 2
    • Windows host boundary isolation and read-only /mnt/c/ path protection
    • Automated syntax linting gates (bash -n, shellcheck, jq empty, py_compile)
    • Lifecycle hook execution and error telemetry capture
  2. Multi-Agent Skill Symlink Sync (scripts/sync_agent_skills.sh):
    • Universal Agent standard relative symlinks (.agents/skills/)
    • Google Antigravity runtime symlinks (~/.gemini/config/skills/)
  3. Settings JSON Configuration Schema:
    • Syntax validation of .claude/settings.json via jq empty

Safety Classification

  • Tier 2 (Controlled System Operations): Executes non-destructive synthetic probes and verification test suites in isolated test environments.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.