gateguard

Skill: gateguard

Context

GateGuard is a pre-edit quality gate that forces agents to investigate a file before making changes. It ensures agents understand:

• Which files import or depend on the target file
• What data schemas, interfaces, and type definitions the file exports
• Whether the user's instructions impose scope constraints

This is the manual/skill layer of the 3-layer enforcement model. The hook layer (PreToolUse/BeforeTool) provides automatic enforcement on Claude CLI and Gemini CLI. This skill covers manual invocation and Antigravity (where hooks don't fire).

Enforcement Coverage

Platform	Hook	Prompt	Skill
Claude Code CLI	✅ PreToolUse	✅	✅
Claude Desktop App	✅* (bundled CLI)	✅	✅
Gemini CLI	✅ BeforeTool	✅	✅
Antigravity	❌	✅	✅

* Claude Desktop App: documented by Anthropic but workspace testing (2026-05) observed intermittent behavior.

When to Use

• Before editing any TypeScript, JSON, or configuration file that may be imported by other files
• When a user request involves modifying shared utilities, type definitions, or script infrastructure
• When the agent is unsure whether a change is safe to make without investigating dependencies
• As a pre-check before any architectural modification to workspace scripts, helpers, or registries

Output Format

[GATEGUARD] Pre-edit investigation complete
  Target: <file-path>
  Importers: <list of importing files, or "none">
  Exported schemas: <list of interfaces/types/classes, or "none">
  Scope constraints: <user-imposed constraints, or "none">
  Risk level: low | medium | high
  Recommendation: proceed | proceed with caution | escalate to architect

Execution Steps

When invoked (via /gateguard <file-path> or trigger match):

1. Receive target file path from arguments or user input
2. Search for importers: Run grep for import patterns referencing the target file's basename

   grep -r "from ['\"].*<basename>['\"]" --include="*.ts" --include="*.js" --include="*.json" --include="*.md" . | grep -v node_modules | grep -v ".git"
   

3. Extract schemas: Scan the target file for interface, type, enum, class exports
4. Summarize findings (1-3 sentences):
   • List importers (if any) and their relationship to the target
   • Note any exported schemas that downstream consumers depend on
   • Flag any scope constraints from user instructions
5. Present to agent: Output summary as [GATEGUARD] Pre-edit investigation complete with findings

Agent Behavior After Investigation

After completing the GateGuard investigation, the agent may proceed with the edit. If importers were found, the agent should verify that changes won't break downstream consumers.

Related Skills

• CONSTITUTION.md §11.2 — GateGuard specification
• scripts/hooks/gateguard-fact-force.ts — Hook layer implementation
• ADR-0021 — Platform Settings Parity Policy