security-pentest-planner — Claude Code skill: pentest plan builder
security-pentest-planner builds a penetration test plan for web applications: analyzes code, API routes, authentication and infrastructure configs, generates pentest-plan.md with concrete test cases.
MIT 297 installs ★ 440 updated 2026-10-20
npx skills add OneWave-AI/claude-skills --skill security-pentest-planner -a claude-codeWhat security-pentest-planner does
The skill reads the codebase with Glob, Grep and Read tools, searches for endpoints, authentication implementation, data storage, file uploads, third-party integrations and security settings. Matches findings against OWASP Top 10 (2021), OWASP API Security Top 10 (2023), CWE Top 25 and compliance frameworks like PCI DSS, HIPAA, GDPR, SOC 2. Flags missing controls, vulnerable dependency versions, hardcoded secrets (existence only), insecure defaults and unvalidated data flows.
Before generating the plan, the skill asks for written authorization to test. Without confirmation, it produces no plan. Writes pentest-plan.md to the project root with 20 sections: test-case tables by category, schedule, tooling, risk assessment methodology, and Rules of Engagement. Every test case references actual file paths, function names, line numbers, endpoint routes. The summary reports endpoint count, test-case count by category, top 5 risk areas and critical issues found during reconnaissance.
What it helps with
- draft a pentest plan for a web application with real test cases
- locate API endpoints, auth modules and security configs in a codebase
- map findings to OWASP Top 10, CWE Top 25 and compliance requirements
- document hardcoded secrets, vulnerable dependencies and missing controls
- prepare a plan with schedule, methodology and testing rules
Example prompts for the agent
- Build a pentest plan for this application, I have written authorization from the owner
- Find all API endpoints and analyze them for OWASP API Top 10 vulnerabilities
- Generate
pentest-plan.mdwith test cases for authentication and authorization
How to install security-pentest-planner
1Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill security-pentest-planner -a claude-code2Manually: put the files in place
~/.claude/skills/security-pentest-planner/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/security-pentest-planner/
1Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill security-pentest-planner -a codex2Manually: put the files in place
.agents/skills/security-pentest-planner/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill security-pentest-planner -a cursor2Manually: put the files in place
.cursor/skills/security-pentest-planner/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install OneWave-AI/claude-skills security-pentest-planner2Manually: put the files in place
.github/skills/security-pentest-planner/SKILL.md1Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill security-pentest-planner -a gemini-cli2Platform CLI
gemini skills install https://github.com/OneWave-AI/claude-skills --path security-pentest-plannerManually: put the files in place
.devin/skills/security-pentest-planner/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add OneWave-AI/claude-skills --skill security-pentest-planner -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about security-pentest-planner
How to install security-pentest-planner in Claude Code?
Run npx skills add OneWave-AI/claude-skills --skill security-pentest-planner -a claude-code, the skill will appear in ~/.claude/skills/security-pentest-planner/SKILL.md.
What does security-pentest-planner do?
It analyzes code, APIs, authentication and infrastructure, maps findings to OWASP Top 10 and compliance requirements, then generates a detailed pentest plan with test cases, schedule and methodology. It requires written authorization before producing any plan.
Can I use security-pentest-planner without system owner permission?
No. The skill asks for written authorization confirmation before generating a plan. Unauthorized access to computer systems violates CFAA, UK Computer Misuse Act and similar laws.
Source text
SKILL.md · MIT
--- name: security-pentest-planner description: Plans security penetration tests for web applications. Analyzes codebase, API routes, auth implementation, and infrastructure config to generate comprehensive pentest plans. For authorized testing only. tools: Read, Glob, Grep, Bash model: inherit --- # Security Penetration Test Planner Act as a senior application security engineer. Analyze a target web application's codebase, API surface, authentication, and infrastructure, then produce a comprehensive `pentest-plan.md` tailored to that specific application. ## Authorization Disclaimer (Required) This skill is for authorized security testing only. Before generating any plan, confirm the user has written authorization. If they have not confirmed, explain that authorization is required and do not produce an offensive plan. Always embed this disclaimer at the top of every generated plan: > This penetration test plan is produced for authorized security assessments only. All testing activities described herein must be performed with explicit written authorization from the system owner. Unauthorized access to computer systems is illegal under the Computer Fraud and Abuse Act (CFAA), the UK Computer Misuse Act, and equivalent laws in other jurisdictions. The author of this plan assumes no liability for misuse. ## Contents - `references/recon-commands.md` -- Phase 1 Glob/Grep search catalog and the full list of data points to collect. - `references/plan-template.md` -- The complete 20-section `pentest-plan.md` output structure with all test-case tables, schedule, tooling, deliverables, risk methodology, and rules of engagement. ## Workflow 1. **Confirm authorization.** Ask the user to confirm written authorization to test the target. Proceed only on confirmation. 2. **Run reconnaissance.** Execute every search in `references/recon-commands.md` systematically. Use Glob to find files, Read to examine them, Grep to find patterns. Cover technology stack, API routes, auth and authorization, data storage, file uploads, third-party integrations, security configuration, and infrastructure. Capture the full data-point checklist in that reference. Record environment variable names only, never values. 3. **Analyze findings.** Cross-reference against OWASP Top 10 (2021), OWASP API Security Top 10 (2023), CWE Top 25, SANS Top 25, and applicable compliance frameworks (PCI DSS, HIPAA, GDPR, SOC 2). Identify missing controls, inconsistent protection across endpoints, vulnerable dependency versions, hardcoded secrets (existence only), insecure defaults, abusable business logic, and unvalidated data flows. Note controls that ARE present as positive findings. 4. **Generate the plan.** Write `pentest-plan.md` in the project root following the exact structure in `references/plan-template.md`. Populate every section with specifics from reconnaissance: actual file paths, function names, line numbers, endpoint paths, and configurations. Reference the real technology stack in the tools section and adjust the schedule to application complexity. Produce no generic boilerplate. 5. **Summarize.** Report to the user: total endpoints discovered, test-case counts by category, top 5 areas of highest concern, recommended immediate actions before the pentest begins, and any critical issues found during reconnaissance. ## Constraints - Never include actual secret values; note existence and location only. - Never execute attacks. This skill generates plans, not exploits. - Always embed the authorization disclaimer at the top of the plan. - Be specific. Every test case must reference real code, endpoints, or config in the target. - Weight severity by business context (a payment endpoint outranks a public blog comment). - Output the plan as `pentest-plan.md` in the project root, comprehensive enough that a qualified tester could execute from this document alone.
© OneWave-AI · MIT · OneWave-AI/claude-skills