oscar-chw/agent-harness
One standard-library install of shared rules, memory and retrieval for Claude Code, Codex, Copilot, Cursor, Gemini CLI and two more AI coding tools (as far as each supports them), plus a fail-closed command guard for Claude Code. The guard
stack: Python
ARCHETYPE
Skill Collector
Ten-plus skills loaded on demand. A procedural-knowledge library.
Copy this rig
$ git clone --depth 1 https://github.com/oscar-chw/agent-harness MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit oscar-chw/agent-harness/content/skills/clear-docs .claude/skills/clear-docs $ npx degit oscar-chw/agent-harness/content/skills/data-job-trial-run .claude/skills/data-job-trial-run $ npx degit oscar-chw/agent-harness/content/skills/debug-systematically .claude/skills/debug-systematically $ npx degit oscar-chw/agent-harness/content/skills/production-code .claude/skills/production-code $ npx degit oscar-chw/agent-harness/content/skills/research-paper .claude/skills/research-paper $ npx degit oscar-chw/agent-harness/content/skills/session-memory .claude/skills/session-memory $ npx degit oscar-chw/agent-harness/content/skills/standards .claude/skills/standards $ npx degit oscar-chw/agent-harness/content/skills/verify-before-done .claude/skills/verify-before-done $ npx degit oscar-chw/agent-harness/content/skills/work-loop .claude/skills/work-loop $ npx degit oscar-chw/agent-harness/content/skills/write-tests-first .claude/skills/write-tests-first
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} Skills (10)
Similar rigs
humanplane/homunculus
A Claude Code plugin that watches how you work, learns your patterns, and evolves itself to help you better.
Minimalist 493 tok · · Q0
princelundgren/prince-skills
Rules, skills, agents and hooks that make coding agents check facts, own mistakes and pressure-test plans. For Claude Code, Qwen Code and Kiro CLI.
Minimalist 772 tok · · Q0
zkoekl/drift-guard
Catch and prevent AI coding agents drifting from your plan and instructions (Cursor-first).
Automator 2.3k tok · · Q23
alexeyban/EKOS
EKOS is an AI-native platform that continuously reconstructs, compiles, stores and serves enterprise knowledge.
Pragmatist 19.2k tok · · Q10