/report-writing
`[Vulnerability] in [Component] Enables [Impact]`
$ npx -y skills add H-mmer/pentest-agents --skill report-writing --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
- Slash command
/report-writing
Context preview
The summary Claude sees to decide when to auto-load this skill.
`[Vulnerability] in [Component] Enables [Impact]`
SKILL.md
report-writing.SKILL.mdReport Writing
Title Formula
`[Vulnerability] in [Component] Enables [Impact]`
Under 15 words. Title Case. Impact-forward. No URLs.
| Bad | Good | |---|---| | XSS in search | Stored XSS in Comment Renderer Executes JavaScript in Admin Context | | IDOR found | IDOR in User API Exposes PII of All Platform Users | | SQL injection | Blind SQL Injection in Search Filter Enables Full Database Extraction |
Structure
1. **Summary** (2-3 sentences): What's broken, what attacker can do, who's affected. 2. **Steps to Reproduce**: Numbered. ONE action per step. Exact URL, method, headers, body. 3. **Impact**: What attacker walks away with. How many users. Business impact. 4. **PoC**: Self-contained file. Screenshots at each step. Video if multi-step. 5. **CVSS 4.0**: Full vector string with justification per metric. 6. **Remediation**: 1-2 sentences. Developer-actionable. Specific fix.
Style Rules
- Human tone, technical but triager-accessible
- Lead with impact, not process
- No padding ("I discovered...", "During my testing...")
- Every sentence adds information
- Never submit without PoC + evidence
Common Mistakes
- Theoretical bugs ("could allow...")
- Screenshots of Burp instead of clear steps
- CVSS overclaiming
- Same bug class on multiple endpoints as one report (should be separate)
- Missing evidence attachment
Bug bounty agent framework for Claude Code, Codex, Gemini, Cursor, Windsurf, Copilot, and OpenClaw — 48 agents, 26 commands, 19 CLI tools, 2 MCP servers, autonomous hunt loops, exploit chain builder.
Repo: H-mmer/pentest-agents
Other skills on pentest-agents.
- /analyze
Analyze recon output with AI to suggest high-value targets and attack strategies. Usage: /analyze <target>
Open skill - /autopilot
Autonomous hunt orchestrator. INSATIABLE in --autonomous mode: enforces an EXHAUSTION CONTRACT (26 canonical hunter classes, surface probe A-I, depth-engine ≥25 attempts/class, wall-clock floor 90 min/target, PRE-COMPLETION GATE before any summary). No early stops, no clarifying
Open skill - /brain
Manage the engagement brain. Subcommands: 'init' to set up, 'brief <target>' for pre-flight, 'status' for overview, 'exhausted [target]' to see dead ends.
Open skill - /chain
Build deep exploit chains — dispatches chain-builder agent. Given bug A, recursively walks the chain graph. Usage: /chain (then describe bug A)
Open skill - /correlate
Run the finding correlation engine to discover attack chains from individual findings.
Open skill - /cost
Show cost tracking and ROI for this engagement.
Open skill

