Audit Skeptic — Security agent for Claude Code
Tries to refute one skill-audit finding through one named verification lens and returns a verdict with its evidence.
How to install Audit Skeptic
Installs to ~/.claude/agents/donald-ada-workinggenius-audit-skeptic.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/donald-ada/workinggenius/HEAD/.claude/agents/audit-skeptic.md -o ~/.claude/agents/donald-ada-workinggenius-audit-skeptic.md Restart Claude Code, or start a new session, for it to be picked up.
What Audit Skeptic does
name: audit-skeptic description: Tries to refute one skill-audit finding through one named verification lens and returns a verdict with its evidence. Spawned by the /skill-audit workflow, three per finding, blind to each other — never to produce findings of its own. tools: Read, Grep, Glob, Bash effort: high color: orange
You are handed one finding from an audit of this plugin's skills, and your job is to refute it. Findings that survive you go into a diff someone will apply to prose th
Alternatives in Security
- Token Auditor — Scans ui/src/ for hardcoded visual values, duplicate components, and shadcn replacement candidates; produces d 79.4k ★
- Audit Verifier — Adversarially verifies one candidate finding from /bug-audit — tries to REFUTE it by reading the code and, whe 335 ★
- Security Verifier — You write and execute PoC tests to PROVE bugs exist 215 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Finding Skeptic
Read-only adversarial verifier for audit findings. Given one finding and one skeptic lens, actively tries to r
Exploitability Skeptic
Phase 4 of the mobile-security-skill pipeline (half of the adversarial exploitation panel). Independently trie
Cook Refute
jeff refute pass. Adversarially test exactly one blocking review, verification, or audit finding before it tri
Vuln Validator
Independently tries to refute candidate vulnerabilities and records a reasoned verdict for each. Must be a dif
Audit Refuter
Dispatched only by the code-audit-frontend member: a read-only adversarial check that tries to refute one holi
Team Reviewer
Adversarial code review and verification. Operates with a single lens per invocation (security, correctness, p
Related Skills
Writing Craft
A blind tested skill for better creative writing. Nine signals three AI judges named after scoring 50 samples
Zp Validate
Adversarially verify a finding by trying to refute it. Usage: /zp-validate
Beat 5
Beat 5 — Review & open PR. Run /review, fix or refute every finding, then /open-pr, then STOP. The operator me