Skip to content
SkillModel-led19,116 tokens in use

bb-local-toolkit

A bug bounty workflow from recon to report that also records where tools, wordlists and clones live on your machine, with rules for dropping weak findings fast.

14,517 stars on the repository MIT Licenseelementalsouls/Claude-BugHunter · skills/bb-local-toolkit
/plugin marketplace add elementalsouls/Claude-BugHunter /plugin install claude-bughunter@elementalsouls
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing bb-local-toolkit costs

179tokens idle
19,116tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

shelldetectednetworkdetectedwrites filesdetectedgitdetectedcredentialsdetected

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Excellent
  • Frontmatter12/15
  • Structure25/25
  • Examples15/15
  • Output15/15
  • Size9/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Needs attention
  • Runs shell commands−5

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 209
  • Writes files−8

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 286
  • Reaches the network−10

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 297
  • Overrides the user's instructions−25

    tells the model to ignore its instructions or to hide what it does from the user

    SKILL.md, line 1100
  • Reads credentials−15

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 1266
  • Runs git commands−5

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 1539

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity100
  • Maintenance84
  • Community68
  • Maturity28
  • Momentum40

bb-local-toolkit is a skill for Claude that covers the whole bug bounty cycle in one document: recon, pre-hunt research, hunting, validation and reporting. Its opening rules push the agent to read the full program scope first, to prove real impact and to drop theoretical findings early.

It adds a tools section with install notes and command patterns for recon utilities, fuzzers, static analysis and scanners, plus checklists for common bug classes such as IDOR, SSRF, OAuth, race conditions, GraphQL and CI/CD pipelines. It is meant for assets covered by a program you have joined.

What You Get

  • A set of working rules such as read the scope first, stop after five empty minutes and switch targets after an hour
  • A method for chaining one finding into related ones, with a table of common chains
  • Checklists for IDOR, SSRF, OAuth, file upload, race conditions, business logic, XSS, SQL injection, GraphQL and more
  • A seven-question gate and four pre-submission gates to run before writing a report
  • Tool notes for subdomain discovery, fuzzing, static analysis and secret scanning

Common Use Cases

  • Starting a new bug bounty target and wanting one document for the whole session
  • Deciding whether a half-confirmed finding is worth a report
  • Looking up how to call a recon tool or wordlist already installed locally

What Makes It Unique

It combines the workflow with a tool map in a single skill, so the agent can pick the right local wordlist or binary without a separate lookup. Its rules are blunt about wasted time: theoretical bugs are dropped, shallow recon loops are cut short and chains are reported as one finding.

Under The Hood

The skill is one long SKILL.md of instructions with no bundled scripts. Following it, the agent runs shell commands for recon and testing, such as subdomain discovery, URL collection, fuzzing, static analysis with semgrep and secret scanning, and it sends requests to the target hosts and to public sources like scope pages and certificate logs. It writes output files and notes to the working folder, clones repositories with git, and its install notes fetch tools from the network. The text also covers credentials: it shows patterns for finding leaked keys and tokens and tells the agent to prove what a found key can reach. One section describes prompt-injection risks in CI/CD workflows, so the file quotes injection phrases as examples. Run it only against assets you are authorized to test.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers