Gatekeeper — Testing & QA agent for Claude Code
Reviews venture evidence and issues a PROCEED, TEST, or PARK recommendation for a stage gate, naming blockers, decisive evidence, and the cheapest next action.
How to install Gatekeeper
Installs to ~/.claude/agents/nacorga-venture-os-gatekeeper.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/nacorga/venture-os/HEAD/.claude/agents/gatekeeper.md -o ~/.claude/agents/nacorga-venture-os-gatekeeper.md Restart Claude Code, or start a new session, for it to be picked up.
What Gatekeeper does
name: gatekeeper description: Reviews venture evidence and issues a PROCEED, TEST, or PARK recommendation for a stage gate, naming blockers, decisive evidence, and the cheapest next action. Use only after research and challenge artifacts exist. tools: Read, Grep, Glob model: opus
You are the venture gatekeeper.
You do not research new facts unless explicitly asked. Judge the evidence already present in the repository against the current gate.
Allowed outcomes:
- PROCEED
- TEST
- PARK
Alternatives in Testing & QA
- Civitai Test Review — Reviews the tests in a feature segment of the main Civitai Next.js app (src/) for whether they would actually 7.2k ★
- Pencil UI Designer — Use this agent when you need Pencil-based design workflows: atomic designer actions, MCP tools, and spec-drive 635 ★
- Lens Reviewer — Use this agent for severity-rated code review with 2-stage protocol (spec compliance first, then code quality) 523 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Sdd Spec Writer
Creates or updates an implementation-independent Next.js feature specification with stable acceptance criteria
Concession Judge
Recommendation Tribunal judge — detects position changes without new evidence (SPEC-192)
Vb Action Analyzer
Analyzes UFT VBScript to extract actions, validations, test data and synchronization, and maps each to a Selen
Data Science Reviewer
Fast Data Science code review focused on research-critical bugs: data leakage, incomplete seeds, temporal leak
Testing Evidence Collector
Screenshot-obsessed, fantasy-allergic QA specialist - Default to finding 3-5 issues, requires visual proof for
Fixme Review Code
Reviews code produced by plan execution. Finds real bugs, gaps, test issues, and inconsistencies. Read-only -
Related Skills
Ask Batch
Ask-batch skill: park piecemeal questions with context, group by decision, ask once per group with proposed de
Proceed With The Recommendation
Execute the agent's prior recommendation list under the 7 Laws — walk in order, route per item, verify, reflec
Prm
Watch a PR (or all your open PRs needing action) with a native Monitor per PR that self-terminates on merge/cl