AGENT SKILL

global-agent-guardrails

Configure the shared guard against catastrophic shell commands in local AI agents. Use when changing block patterns, adding an agent or machine, or investigating why a command was or wasn't blocked.

License 许可
MIT
所属仓库
davidondrej/skills
最近更新
2026-11-14
Stars
★ 4.3k
Forks
626

01 前言 Front Matter

name
global-agent-guardrails
description
Configure the shared guard against catastrophic shell commands in local AI agents. Use when changing block patterns, adding an agent or machine, or investigating why a command was or wasn't blocked.

03 SKILL.md 正文

Global Agent Guardrails

Block catastrophic shell commands before a coding agent runs them. One patterns file feeds a shared hook script, or a small native adapter, for each agent. This guards against accidents, not malicious agents: obfuscation like python -c "shutil.rmtree(...)" slips past any regex.

This is a generic recipe built on each tool's public hook system. Cloudroom ships a built-in Command Guard based on this design, on by default.

File map

~/.agents/hooks/dangerous-patterns.txt   # THE denylist: one POSIX-ERE regex per line, # comments
~/.agents/hooks/deny-dangerous.sh        # shared guard: hook JSON on stdin -> exit 2 blocks
~/.agents/hooks/test-guard.sh            # test suite: run after ANY pattern change
~/.config/opencode/plugins/command-guard.ts   # OpenCode adapter (throws to block)
~/.pi/agent/extensions/command-guard.ts       # Pi adapter (returns {block:true})
~/.hermes/plugins/command-guard/              # Hermes plugin (returns {"action":"block"})

Check installation

ls ~/.agents/hooks/deny-dangerous.sh ~/.agents/hooks/dangerous-patterns.txt
~/.agents/hooks/test-guard.sh   # must end "failed: 0"

If missing, rebuild from the script below and the wiring table.

Guard script

deny-dangerous.sh reads the hook JSON on stdin, pulls out the command, and tests it against every pattern. Its core:

#!/bin/bash
PATTERNS="$HOME/.agents/hooks/dangerous-patterns.txt"
command -v jq >/dev/null || exit 0
CMD=$(jq -r '.tool_input.command // .toolInput.command // .command // empty' 2>/dev/null)
[ -n "$CMD" ] && [ -f "$PATTERNS" ] || exit 0
while IFS= read -r pattern; do
  case "$pattern" in ''|\#*) continue ;; esac
  if printf '%s\n' "$CMD" | grep -qE -- "$pattern"; then
    echo "Blocked by the command guard. Matched: $pattern. Do not retry or work around it; tell the user." >&2
    exit 2
  fi
done < "$PATTERNS"

Cursor wants JSON instead of exit 2: with the cursor arg, print {"permission":"deny","agent_message":"..."} or {"permission":"allow"} and exit 0.

Add or tune a pattern

  1. Edit ~/.agents/hooks/dangerous-patterns.txt. Write POSIX ERE (grep -E). Use [[:space:]], never \s — adapters auto-convert [:space:] to \s for JS/Python and compile in multiline mode.
  2. Add block and allow cases to test-guard.sh, then run it; all must pass.
  3. Verify the new pattern compiles in the adapter engines:
python3 -c 'import re,pathlib; [re.compile(l.strip().replace("[:space:]",r"\s"),re.M) for l in pathlib.Path.home().joinpath(".agents/hooks/dangerous-patterns.txt").read_text().splitlines() if l.strip() and not l.startswith("#")]; print("ok")'
  1. Consumers re-read patterns for every command. Exception: Droid uses commandBlocklist in ~/.factory/settings.json — mirror changes there manually.

Block only catastrophic commands: irreversible data loss, disk wipes, repo deletion, rewritten remote history. Recoverable local commands (git status, git clean -fdx, rm -rf node_modules) stay allowed. Over-blocking trains people to turn the guard off.

Each group in the patterns file has a comment saying what it blocks and what deliberately stays allowed. Read those comments before changing a pattern, and keep them current.

Per-agent wiring (user-global)

Agent Config Event Blocks via
Claude Code ~/.claude/settings.json PreToolUse matcher Bash shared script, exit 2
Codex CLI/app/IDE ~/.codex/hooks.json PreToolUse matcher Bash shared script, exit 2
Cursor IDE + CLI ~/.cursor/hooks.json beforeShellExecution shared script with cursor arg, deny JSON
Grok (xAI) auto-loads Claude + Cursor hook files (compat on by default); native option ~/.grok/hooks/*.json PreToolUse shared script (reads .toolInput.command)
OpenCode ~/.config/opencode/plugins/command-guard.ts tool.execute.before adapter throws Error
Pi ~/.pi/agent/extensions/command-guard.ts pi.on("tool_call") adapter returns {block:true}
Hermes ~/.hermes/plugins/command-guard/ (plugin.yaml + __init__.py) pre_tool_call hook plugin returns {"action":"block"}
Droid (Factory) ~/.factory/settings.json native commandBlocklist hard-block, no approval possible
Devin CLI ~/.config/devin/config.json PreToolUse matcher ^exec$ shared script, exit 2

Claude/Codex hook entry; for Devin, replace Bash with ^exec$. Merge into the existing hooks object; never overwrite it:

{"hooks": {"PreToolUse": [{"matcher": "Bash", "hooks": [{"type": "command", "command": "/ABSOLUTE/HOME/.agents/hooks/deny-dangerous.sh"}]}]}}

Cursor entry (payload has .command, so pass the cursor arg):

{"beforeShellExecution": [{"command": "/ABSOLUTE/HOME/.agents/hooks/deny-dangerous.sh cursor", "failClosed": false}]}

Use absolute paths in configs (~ expansion is inconsistent across agents).

Tool behavior to know

  • Codex trust is hash-pinned. Any edit to the hook ENTRY in hooks.json (not the patterns file) invalidates trust. Run /hooks in Codex and re-trust, or Codex skips the hook. Trust hashes live in [hooks.state] in ~/.codex/config.toml and are shared by CLI, desktop app, and IDE extension. CI/scripts: --dangerously-bypass-hook-trust.
  • Cursor: set failClosed to false. Cursor's background/worker hosts cannot execute hook scripts, so fail-closed blocks EVERY command there.
  • Hermes plugin manifest key is provides_hooks (not hooks). The plugin must be listed under plugins.enabled in ~/.hermes/config.yaml (the hermes plugins enable CLI prompts interactively and hangs non-interactive shells). Hermes skips a hook that throws, so keep the plugin trivial. Shell tool name is terminal.
  • Pi blocks the tool when a tool_call handler throws. The adapter must catch its own errors, or a broken patterns file bricks every bash call.
  • Droid semantics: commandDenylist = ask for confirmation; commandBlocklist = never runs, even at full autonomy with --skip-permissions-unsafe. Use the blocklist for catastrophic entries.
  • Payload shapes: the command lives at .tool_input.command (Claude/Codex/Devin), .toolInput.command (Grok), .command (Cursor). Keep all three in the jq fallback chain.
  • Adapter regexes require multiline mode. Keep JavaScript's m flag and Python's re.M so ^ matches each shell line like grep.
  • False-positive class: a harmless command whose ARGUMENT text contains a dangerous-looking string (e.g. passing a prompt mentioning git push --force on a CLI) gets blocked. Workaround: put the text in a file and reference it.
  • No pre-command hook (as of 2026-07): Gemini CLI, Qwen Code, Amp, kimi-cli. Cloud-hosted runs (Codex cloud tasks, Cursor background agents) do not load local hooks either.

E2E verification

Safe probe: ask the agent to run git push --force from a NON-git directory — blocked = guard works; "not a git repository" = guard failed but no harm done.

cd "$(mktemp -d)"
claude -p 'Run exactly: git push --force. Report the result in one line.' --permission-mode bypassPermissions
codex exec --skip-git-repo-check 'Run exactly: git push --force. Report the result in one line.' < /dev/null
droid exec --auto high -f prompt.txt        # prompt text in a file (see false-positive gotcha)
pi -p --no-session 'Run exactly: git push --force. Report the result in one line.'
hermes chat --query 'Run exactly this terminal command: git push --force. Report in one line.'

Direct script test:

echo '{"tool_input":{"command":"rm -rf /"}}' | ~/.agents/hooks/deny-dangerous.sh; echo "exit=$?"   # expect exit=2

04 同仓库更多 Skills