Works with the AI Agents you already use
audit fortress
A deterministic security audit corps for finding, fixing, and gating vulnerabilities in code and CI pipelines.
$24.90
audit fortress
Example session with this skill installed
Run a security sweep on the current repository and show me the Bounty Board.
- Read your context and instructions
- Compiled the audit fortress
I have completed the security sweep of the repository. The audit corps analyzed the source code and CI configurations, identifying three high-priority findings. These have been recorded in the Bounty Board ledger. ## Bounty Board Audit Summary | ID | Hunter | Exploit Class | Status | Evidence | |---|---|---|---|---| | INJ-001 | Injection Hunter | SQL Injection | Open | `db.execute()` sink in `auth.py:42` via `user_id` | | SEC-004 | Secrets Hound | Hardcoded Token | Open | Plaintext AWS key in `.env.example` | | SUP-012 | Supply-Chain Sentry | CI Over-privilege | Open | `GITHUB_TOKEN` set to `write` in `deploy.yml` | ### Threat Model: Component Boundaries | Component | Trust Boundary | STRIDE Threat | Mitigation | |---|---|---|---| | Web API | External / Public | Information Disclosure | Implement rate limiting on /api/v1 | | DB Connector | Internal / VPC | Tampering | Parameterized queries in `auth.py` | | CI Pipeline | Runner Context | Elevation of Privilege | Pin actions to SHAs, restrict permissions | **Next steps** 1. Run `bastion fix INJ-001` to generate the parameterized query patch and pinning test. 2. Review the `deploy.yml` workflow to restrict `permissions:` to `contents: read`. 3. Execute `bastion gate` once fixes are applied to verify closure in the ledger.
Connects securely to your tools. The creator never sees your data.
About this skill
Security debt accumulates faster than most teams can patch. This skill provides a deterministic security audit corps that finds vulnerabilities in your own code and CI configurations without the noise of generic scanners. It delivers evidence-backed findings including the exact sink, reachable path, and a minimal fix. ## What it does - **Injection analysis** identifies SQL, command, path, and template injection paths via taint analysis. - **Authorization hunting** maps endpoints to roles to detect missing checks and IDOR patterns. - **Secrets detection** locates credentials and tokens while providing specific rotation guidance. - **Supply-chain review** audits dependencies, lockfiles, and CI permissions for confusion patterns. - **Threat modeling** generates STRIDE tables and trust boundaries based on actual source code. - **Regression pinning** creates minimal patches and pinning tests to ensure closed findings stay closed. ## How it works 1. **Sweep** the repository using `bastion sweep` to generate a list of findings with evidence chains. 2. **Model** the attack surface with `bastion threatmodel` to visualize trust boundaries and STRIDE risks. 3. **Fix** identified vulnerabilities using the `bastion fix` command to receive patch guidance and tests. 4. **Gate** the project with `bastion gate` to verify closures and update the monotonic ledger. ## Frameworks & tools Built for Python 3.10+ environments. It analyzes any source language with deep support for Python, JavaScript, TypeScript, and CI configuration files. ## Why this beats prompting it yourself Generic prompts often hallucinate security risks or miss complex taint paths across multiple files. This skill uses a structured registry and the Bounty Board ledger to ensure every finding is backed by an evidence chain and a regression-proof closure gate. ## Use cases - Auditing a repository before a major release or production deployment. - Hardening CI/CD pipelines by scoping tokens and pinning action SHAs. - Preparing a verified audit trail for customers or compliance reviewers. - Detecting and rotating leaked secrets within the development history. ## Known limitations Requires Python 3.10 or higher. While it reads all source code, deep analysis is optimized for Python, JavaScript, and TypeScript.
Requires Python 3.10+. Deepest analysis coverage is restricted to Python, JavaScript, TypeScript, and CI configurations.
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 any payment.
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 28 days ago
Requires
Permissions
Allowed Hosts