Stats
Actions
Tags
From tool-gates
Review and approve pending tool-gates permissions. Lists commands you've been manually approving with counts and suggested glob patterns, lets you multi-select which to permanently allow, and writes rules to settings.json.
How this skill is triggered — by the user, by Claude, or both
Slash command
/tool-gates:reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
1. Run `tool-gates pending list $ARGUMENTS` to get pending approvals.
Run tool-gates pending list $ARGUMENTS to get pending approvals.
--all: shows all projectsIf no pending approvals, tell the user and stop.
Present as a numbered list using the first suggested pattern for each:
1. [ ] cargo build --release (12x) -> cargo build:*
2. [ ] npm install (8x) -> npm install:*
3. [ ] git push origin main (3x) -> git push:*
If multiple patterns are suggested, show the most specific one. Mention alternatives briefly (e.g., "or broader: cargo:*").
Ask the user which to approve (e.g., "1, 3" or "all") and at what scope:
.claude/settings.local.json).claude/settings.json)~/.claude/settings.json)For each selection, run tool-gates approve '<pattern>' -s <scope>.
Show final rules with tool-gates rules list.
git:* allows force-push, rm:* allows recursive delete)git push:*) over broad ones (git:*) unless the user explicitly wants broadnpx claudepluginhub camjac251/tool-gates --plugin tool-gatesGenerates PermissionRequest hooks that auto-approve safe operations, auto-deny dangerous ones, and tailor rules to detected project stack. Safer alternative to --dangerouslySkipPermissions for manual permission mode.
Reviews, summarizes, and edits opencode permission configs including always-allow lists, skill permissions, and wildcard patterns. Recommends safe read-only commands for auto-approval and audits for security.
Analyzes permission denial patterns and generates optimized alwaysAllow and alwaysDeny rules to reduce prompt fatigue.