bb-methodology

Orchestrates bug bounty hunting sessions using a 5-phase workflow and critical thinking framework.

Updated May 29, 2026
One-click install
npx skills add https://github.com/hhjkjkjk/Claude-skills --skill bb-methodology-hhjkjkjk
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: bb-methodology
Source: https://github.com/hhjkjkjk/Claude-skills/tree/main/skills/bb-methodology
Command: npx skills add https://github.com/hhjkjkjk/Claude-skills --skill bb-methodology-hhjkjkjk

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Bug bounty hunters often waste sessions wandering without a goal, report hygiene findings that get rejected as N/A, or get stuck in rabbit holes. This Skill provides a master orchestrator that structures every hunting session around a 5-phase non-linear workflow (Recon, Mapping, Discovery, Prove & Escalate, Validate & Report) combined with a critical thinking framework covering developer psychology, anomaly detection, and What-If experiments. ## Core Features & Use Cases - Engagement Mode Confirmation: Distinguishes bug bounty, red team, pentest, and internal audit engagements so only impact-demonstrated findings get reported on bounty platforms. - 5-Phase Non-Linear Workflow: Routes between Recon, Mapping, Vulnerability Discovery, Escalation, and Reporting with explicit decision tables, tool routing (subfinder, nuclei, ffuf, dalfox, interactsh), and 20-minute rotation rules. - False-Positive Prevention Discipline: Enforces marker discipline, body-diff verification, statistical sampling for timing claims, and a shell-loop ban to eliminate retracted findings. - Use Case: Starting a session on a new HackerOne program, invoke this Skill to define a daily goal (e.g., IDOR on the orders API for account takeover), follow the phase checklists, escalate a confirmed IDOR to ATO, and pass the 7-question validation gate before writing the report. ## Quick Start Start a bug bounty hunting session on target example.com and ask what phase to begin with and which vulnerability classes to focus on first.

Frequently Asked Questions about bb-methodology

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I start a bug bounty hunting session on a new target?▼

Begin with Phase 0: define a specific goal like targeting a feature for a CIA impact, select one or two vulnerability classes such as IDOR or XSS, then run a wide recon sweep with subdomain enumeration and HTTP probing before mapping the application.

What is the difference between bug bounty and red team findings?▼

Bug bounty engagements accept only impact-demonstrated bugs with a full chain to attacker-attainable harm, while red team engagements treat hygiene findings, recon data, and defensive observations as valid deliverables. Confirming the mode before testing prevents rejected reports.

How do I avoid false positives when testing for reflection or bypasses?▼

Use unique random markers of 8 or more characters and check the baseline response for natural occurrences before claiming reflection. For bypass claims, diff the response body against the baseline since a 200 status with an identical body is not a bypass.

When should I move on from a parameter or endpoint that is not working?▼

Apply the 20-minute rotation rule: if you are not making progress after 20 minutes on an endpoint, rotate to the next endpoint, subdomain, or vulnerability class. Cap any single parameter at 45 minutes to avoid rabbit holes.

How do I escalate a low-impact bug into a critical finding?▼

Follow the escalation decision tree: chain XSS to session theft or forced email change for account takeover, escalate SSRF to cloud metadata extraction, and automate IDOR data scraping to demonstrate scale. Minimize attack prerequisites to zero or one click.

Why do timing-based user enumeration claims get rejected?▼

Single timing outliers are usually network jitter. Valid claims require at least 10 interleaved trials per group with the suspect group mean at least two standard deviations above the control group mean, otherwise the signal is not statistically reliable.