verification-gate

A verification process for checking a code change or pull request against its acceptance criteria and known regressions. It produces evidence from reproduction steps, tests, linting, type checks, and CI checks before giving a pass, fail, or blocked result.

In plain words
What is it for?
Use it after a fixer creates a commit or pull request to run focused and broader checks, inspect GitHub checks, record results, and issue a final verdict.
Why use it?
It separates independent validation from implementation and makes it clear whether a change works, what failed, and what still needs human review.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/wellkilo/repopilot/verification-gate
Any agent
npx skills add wellkilo/RepoPilot --skill verification-gate
Clone the repo
git clone --depth 1 https://github.com/wellkilo/RepoPilot

Made for: Claude Code, Codex.

Per session 38 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 440 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00038 $0.00440
Opus 5 $0.00019 $0.00220
Sonnet 5 $0.00008 $0.00088
Haiku 4.5 $0.00004 $0.00044

Measured yesterday against content hash be937c490ba3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

verification-gate scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/verification-gate/SKILL.md · 64 lines

What it actually says

Verification Gate

You are the Verifier Worker. Independently determine whether the patch satisfies the acceptance criteria without introducing a known regression.

Inputs

  • RepoPilot runId
  • Base and patched revisions
  • Locator reproduction evidence
  • Fixer change summary and pull request
  • Triage acceptance criteria

Outputs

  • Before/after reproduction result.
  • Focused test result.
  • Relevant broader test, lint, and type-check results.
  • GitHub pull request check status.
  • Verdict: PASS, FAIL, or BLOCKED.
  • Residual risk and any required human review.

Append each command result as ci_result or tool_result evidence.

Invocation Conditions

Use after Fixer produces a commit. Repeat only for a new commit SHA.

Call repopilot_start_step before execution with a stable attempt-specific idempotencyKey. Call repopilot_finish_step exactly once with the final succeeded, failed, blocked, skipped, or cancelled outcome.

Dependencies

  • Repository-local test tools
  • github_list_pull_request_checks
  • repopilot_append_evidence

Failure Handling

  • Distinguish changed-code failures from pre-existing failures.
  • Retry only tests documented as flaky and record every attempt.
  • If CI is pending, return BLOCKED; do not infer success.
  • If the target behavior cannot be observed, return BLOCKED.

Permission and Safety Boundary

  • Read-only verification.
  • Do not amend the patch, merge the pull request, or dismiss failing checks.
  • A green result is not merge approval.

Reuse Value

The evidence-based verdict contract is portable to dependency upgrades, test additions, documentation builds, and security remediations.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 64 lines · 38 tokens per session scan A be937c490ba3

Subscribe to this mod's changes

verification-gate is a skill published in the GitHub repository wellkilo/RepoPilot (2 stars, last pushed 8d ago), licensed Apache-2.0. It adds 38 tokens to every session and 440 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.