capture0x/YeepForge / .claude/skills/yeepforge-pentest
yeepforge-pentest
VerifiedRun a full web-application penetration test by driving YeepForge's MCP tools. Use when the user gives a target URL (with authorization) and wants reconnaissance, OWASP Top 10 testing, exploitation, and a professional report. Requires the 'yeepforge' MCP server to be connected.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.5 KB |
SKILL.md
View on GitHub---
name: yeepforge-pentest
description: "Run a full web-application penetration test by driving YeepForge's MCP tools. Use when the user gives a target URL (with authorization) and wants reconnaissance, OWASP Top 10 testing, exploitation, and a professional report. Requires the 'yeepforge' MCP server to be connected."
---
# YeepForge - Autonomous Web Application Pentest
This skill orchestrates a complete web-app engagement through the **`yeepforge`
MCP server**. Every tool below is exposed as `mcp__yeepforge__<name>`. YeepForge
injects the engagement target/cookies/proxy into each call, so you set the target
**once** and never repeat it.
## Rule 0 - Authorization gate (do this first, every time)
Do **not** run a single scan tool until the user has confirmed they own the
target or have **written authorization** to test it. If scope is unclear, ask.
State the target back to the user and get a yes before proceeding.
## Phase 1 - Set engagement
Call `set_engagement` with the `target_url` (and `cookies`, `auth_token`,
`proxy`, `username`/`password`, `scope` if provided). This is the only place the
target is entered.
## Phase 2 - Recon & mapping (always run before testing)
1. `recon_target` - HTTP headers, server/tech fingerprint (curl + whatweb).
2. `crawl_target` - discover pages, forms, JS endpoints, API/GraphQL routes.
Discovered endpoints feed every later test.
3. `directory_scan` - hidden dirs/files (gobuster/ffuf).
4. `find_sensitive_files` - `.env`, `.git`, backups, actuator/debug endpoints.
5. `fuzz_parameters` - hidden GET/POST params to widen the injection surface.
Look up any interesting software versions with `web_search` (`search_type:"cve"`).
## Phase 3 - OWASP Top 10 testing
Pick tests based on what recon found; don't blindly fire everything.
| Class | Tool |
|-------|------|
| SQL injection | `test_sql_injection` |
| XSS | `test_xss` |
| SSRF | `test_ssrf` |
| Path traversal | `test_path_traversal` |
| Command injection | `test_command_injection` |
| SSTI | `test_ssti` |
| XXE | `test_xxe` |
| CORS misconfig | `test_cors` |
| Open redirect | `test_open_redirect` |
| File upload | `test_file_upload` |
| GraphQL | `test_graphql` |
| Auth / JWT / default creds | `test_authentication` |
| Security headers | `check_security_headers` |
| TLS/SSL | `check_ssl_tls` |
| Broad scan | `scan_vulnerabilities` (nikto), `nuclei_scan` |
For manual/edge-case payloads, pull from the vendored reference skills:
`security-payloads` (XSS/XXE/SSTI/upload), `security-fuzzing` (SQLi/NoSQL/cmd),
`security-patterns` (secret/PII regexes for triaging responses).
## Phase 4 - Confirm & exploit
- The moment a vuln is **confirmed** (not suspected), call `report_finding`
with title, severity, OWASP category, evidence, and remediation.
- To demonstrate impact on a confirmed finding, use `exploit_vulnerability`
(SQLi→dump, RCE→shell, SSRF→metadata, traversal→file read). Only after
confirmation, and only within scope - no destructive actions.
## Phase 5 - Report
Call `generate_report` for a CVSS-scored HTML report, then `finish_assessment`
with a short summary of coverage and overall risk.
## Source-code available?
If the user also provides a source path, run the companion `yeepforge-sast`
skill (or call `analyze_source_code`) to confirm dynamic findings at the code
level and catch what black-box testing misses.
## Guardrails
- Authorized targets only. Respect scope and rate limits; avoid DoS.
- Report **confirmed** findings only - no speculation in the report.
- Prefer targeted tests over noisy broad scans when the target is fragile.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |