Works with the AI Agents you already use
Audit Gate Pro — Defensive Repository Security Audit
A procedural security audit gate that identifies blocking vulnerabilities with concrete exploitation scenarios.
$13
Audit Gate Pro — Defensive Repository Security Audit
Example session with this skill installed
Run a security audit on the current repository before the v1.2.0 release. Focus on secrets and API vulnerabilities.
- Read your context and instructions
- Compiled the audit gate pro
The security audit for core-api-service is complete. I have reviewed the source code, environment configurations, and dependency manifests.
Audit: core-api-service (2024-05-22)
| Gate | Verdict | Blocking findings |
|---|---|---|
| 1 Secrets | FAIL | 1 |
| 2 Dependencies | PASS | 0 |
| 3 Static analysis | FAIL | 1 |
| 4 Configuration | PASS | 0 |
| 5 PII / GDPR | PASS | 0 |
| 6 Architecture | PASS | 0 |
Blocking findings (CRITICAL/HIGH)
src/config/auth.ts:14 CRITICAL JWT with algorithm: "none"
Exploit: Attacker forges tokens without a key, leading to total authentication bypass.
Fix: Explicitly set algorithms: ['HS256'] in the verification options.
Value: [REDACTED...none]
src/controllers/userController.ts:42 HIGH Mass assignment (...req.body)
Exploit: Attacker sends extra JSON fields like isAdmin to escalate privileges during profile updates.
Fix: Use a DTO or whitelist specific fields for the update operation.
Backlog (non-blocking): 1 MEDIUM, 2 LOW, 0 INFO
src/lib/db.ts:89MEDIUM: Secret compared with===instead of constant-time comparison.package.jsonLOW: Unpinned dependency range forlodash.
Verdict: BLOCKED
Next steps
- Fix the JWT verification logic in
auth.tsto prevent token forgery. - Implement a whitelist for the user update endpoint to prevent mass assignment.
- Rotate the
JWT_SECRETin production as the "none" algorithm usage may have exposed it.
Connects securely to your tools. The creator never sees your data.
About this skill
Release cycles often skip thorough security reviews due to time pressure or expensive third-party audits. This skill implements a procedural defensive audit protocol to catch critical vulnerabilities before they reach production.
What it does
- Secrets scanning identifies hardcoded private keys, JWTs with weak signatures, and recognizable API key formats.
- Dependency auditing parses package manager manifests to detect unpinned versions, missing lockfiles, and known CVEs.
- Static analysis reviews code for injection vectors, IDOR vulnerabilities, mass assignment, and path traversal.
- Configuration checking flags environment variables committed to version control and insecure
.gitignoresettings. - PII/GDPR analysis detects excessive data exposure in API responses and sensitive fields leaked in logs.
- Architectural review identifies broken dependency inversion where domain logic couples to infrastructure vendors.
How it works
- Repository discovery via read-only file exploration and
git ls-filesto identify high-risk surfaces. - Procedural gate checks (G1-G7) executed sequentially, covering secrets, dependencies, code, and config.
- Severity classification applying a strict blocking policy for CRITICAL and HIGH findings while backlogging lower risks.
- Report generation including specific
file:linelocations and concrete exploitation scenarios for every finding.
Frameworks & tools
Works with any repository using npm, yarn, or pnpm. Analyzes codebases written in TypeScript, JavaScript, and related web frameworks, focusing on REST/GraphQL APIs and server-side logic.
Why this beats prompting it yourself
Generic prompts often return vague "best practice" advice or hallucinate vulnerabilities. This skill follows a strict "no exploit, no finding" rule and enforces a specific blocking policy to prevent alert fatigue.
Use cases
- Pre-merge gate for critical pull requests to ensure no secrets or SQL injection vectors are introduced.
- Legacy takeover to quickly map the security posture and technical debt of an unfamiliar codebase.
- Release validation to verify dependency health and configuration hygiene before deploying to production.
Known limitations
Does not perform dynamic analysis or network-based penetration testing. The IDOR check requires manual verification against router-level guards to minimize false positives.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.
Read-only analysis only. No dynamic execution or network scanning. IDOR detection requires manual confirmation of auth guards. Limited to JS/TS ecosystems for dependency parsing.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 20 days ago
- Software
Git