Op Verifier — Development agent for Claude Code
Fresh-context adversarial verifier, dispatched by the operator after implementation work — independently tries to refute the claimed outcome and returns CONFIRMED or REFUTED with evidence.
How to install Op Verifier
Installs to ~/.claude/agents/betmoar-cc-operator-plugin-op-verifier.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/betmoar/cc-operator-plugin/HEAD/agents/op-verifier.md -o ~/.claude/agents/betmoar-cc-operator-plugin-op-verifier.md Restart Claude Code, or start a new session, for it to be picked up.
What Op Verifier does
name: op-verifier description: Fresh-context adversarial verifier, dispatched by the operator after implementation work — independently tries to refute the claimed outcome and returns CONFIRMED or REFUTED with evidence. model: opus effort: medium tools: Read, Grep, Glob, Bash disallowedTools: Write, Edit, NotebookEdit
You verify exactly one claim per dispatch ("X was implemented and works") with fresh eyes. Assume it is broken until evidence you produce yourself says otherwise. Re-run th
Alternatives in Development
- Finding Checker — Blind adversarial checker — given ONLY a finding artifact and its evidence (never the author's reasoning), tri 348 ★
- Cold Verifier — Phase 11 cold verification agent that independently re-verifies CRITICAL and HIGH findings with zero prior con 125 ★
- Tech Debt Verifier — Adjudicates a single candidate tech-debt finding produced by the tech-debt-assessor hunter 107 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Refute Verifier
Clean-context skeptic that tries to REFUTE each draft finding or claim it is given — verdicts CONFIRMED / REFU
X Review Verifier
Adversarial verifier for review findings. Given a finding and the diff, tries to refute it. Returns refuted wh
Claim Auditor
Adversarial verifier for proposed genealogical matches. Tries to break each proposed match with evidence, kill
Self Review Verifier
Read-only adversarial verifier for the self-review loop. Takes one or more candidate findings plus the scope f
Victor
Adversarial verifier. Tries to refute one raised finding; uncertainty refutes, only a shown broken path confir
Aidd Adversarial Verifier
Tries to REFUTE one AIDD finding; only findings surviving a motivated skeptic block delivery.
Related Skills
Lane Verify
Out-of-domain verification lane. Independently REFUTE other lanes' done-claims, defaulting to refuted when unc
Reviewer Verify
You are the adversarial verifier for the candidate pull-request findings raised against one file. Your only jo
Evil Mode
EVIL mode — adversarial sweep with scored findings, adversarial verification, and loop-until-dry. The upgrade