AI Gate — AI skill for Claude Code
AI verification for G4 — eval suite against baseline, red-team status, guardrail verification, drift check.
How to install AI Gate
Installs to ~/.claude/commands/toddgbenson-keel-ai-gate.md
mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/ToddGBenson/keel/HEAD/.claude/commands/ai-gate.md -o ~/.claude/commands/toddgbenson-keel-ai-gate.md Restart Claude Code, or start a new session, for it to be picked up.
What AI Gate does
description: AI verification for G4 — eval suite against baseline, red-team status, guardrail verification, drift check argument-hint: allowed-tools: Read, Grep, Glob, Write, Edit, Bash, WebSearch, Task
G4 — AI Verification: `$ARGUMENTS`
Delegate to the `ai-risk-officer` agent. Per `docs/12-ai-feature-governance.md`.
Applies when
The change is **AI-relevant** per `docs/10-definitions.md`: model call added or changed; prompt, system instruction, or generation
Alternatives in AI
- Datafog Python — Offline PII firewall for AI agents and LLM apps: fast local detection and redaction, Claude Code hook, LiteLLM 69 ★
- Moyu — claude skill install --url https://github.com/uucz/moyu --skill moyu Anti-over-engineering skill that teaches 34 ★
- Fable Status — Report fable-mode posture — output style, model/effort, hook telemetry drift counts 33 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Red Line Skill
A skill for AI agents (Claude, ChatGPT, etc.): clarify scope before planning, lay out an explicit end-to-end p
Driftproof
Continuous, model-version-bound verification of agent skills: run a skill's eval suite with and without the sk
Agent Eval Workbench
MLflow-based evaluation kit for AI agents: golden-task runner, calibrated LLM judge, list-price cost accountin
Agentic Review Gate
Governed, auditable AI review platform: the review and approval work a team does by hand, as a webhook-trigger
Performance Marketing Eval
15 performance marketing tasks for evaluating AI agents, with simulated tools, a shared grader, and baseline t
EvalPlan
Phase 3 (+ai) — build golden / adversarial / regression eval sets, graders, thresholds, baseline. PT - plano d
Related Agents
Offensive Security Engineer
Red-team the verification engine. Use to hunt false SAFE verdicts, shadow-tree isolation escapes, credential l
AI Risk Officer
Owns AI-specific risk in both directions — AI features the team ships, and the AI agents the team builds with.
Gate Auditor
Adversarially audit a test suite or release gate by trying to make every green check go red. Use before trusti
