Completion Gate — Development agent for Claude Code
Lie detector for agent claims — validates every completion claim has actual evidence.
How to install Completion Gate
Installs to ~/.claude/agents/rune-kit-rune-completion-gate.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/Rune-kit/rune/HEAD/agents/completion-gate.md -o ~/.claude/agents/rune-kit-rune-completion-gate.md Restart Claude Code, or start a new session, for it to be picked up.
What Completion Gate does
name: completion-gate description: "Lie detector for agent claims — validates every completion claim has actual evidence. Default-FAIL mindset. Use as final gate before merge/commit." model: sonnet subagent_type: general-purpose
You are the **completion-gate** skill — Rune's claims validator.
Quick Reference
**Workflow:**
- Extract all completion claims from agent output ("tests pass", "build succeeds", "fixed", etc.)
- Stub detection: scan new files for TODO/NotImplementedError/
Alternatives in Development
- Review Fact Checker — Verify every factual claim in a review draft against vault sources 4.6k ★
- PR Check Watcher — Watches a GitHub PR's CI checks to completion via gh pr checks --watch and reports a concise pass/fail summary 289 ★
- Writer Critic — Manuscript critic that reviews paper drafts for structure, claims-evidence alignment, identification fidelity 250 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Vouch Verifier
Read-only verifier one tier below the producer. Checks a worker's CLAIM lines against real evidence and return
Checklist Verifier
Verifies a single claim from the verification checklist against the actual source code. Reports PASS, FAIL, or
Deepreview Validator
Cross-validates code review findings by checking claims against actual source code. Part of the deepreview pip
Skeptic Verifier
Adversarial refuter for load-bearing claims. Recomputes empirical numbers from raw sources and actively tries
Coder Teammate
Code implementation teammate. Works in Agent Teams mode, communicates directly with evaluator-teammate. Claims
Etrnl Adversary
Use this agent when ETRNL needs a read-only adversarial challenge of a plan, diff, or completion claim. Exampl
Related Skills
Final Validation
Use before a push, merge, release, or completion claim for any feature, refactor, UI, documentation, or depend
Verification Before Completion
Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires runn
SecondLook
BCC detector with Claude as a Grad-CAM trust layer - attention check for the doctor's final call.
