Skip to content
← Tools / Skills
ToolTransilience AI/10 trustMetrics-checked

Transilience Community Tools

Open-source Claude Code skills, agents, and slash commands for pentesting and bug bounty hunting.

Visit official site →

About

transilienceai/communitytools is an open-source collection of Claude Code skills, agents, and slash commands purpose-built for AI-powered penetration testing, bug bounty hunting, and security research. It organizes 26 skills across vulnerability testing (10), reconnaissance (3), specialized workflows (5), platform integrations (2), and tooling (6), plus tool integrations and MCP servers, with a multi-agent execution flow for orchestrated security testing.

The repo is MIT-licensed with around 563 stars since its November 2025 creation and includes benchmarks, documentation, and a responsible-disclosure policy. It is one of the few skills collections targeting offensive-security workflows specifically, rather than general coding assistance.

Limitations: this is dual-use tooling — powerful for legitimate security research and bug bounties, but the same skills could aid malicious use, so install and use only within authorized scopes and disclosed programs. Skills run with the agent's permissions; review contents before use. Coverage is security-specific, not a general-purpose skills library.

What it can do

  • 26 security skills: vuln testing, recon, tooling
  • Multi-agent execution flow for orchestrated testing
  • MCP servers and slash-command integrations

Setup

Clone https://github.com/transilienceai/communitytools and follow the local or Docker setup in the README; run skills from Claude Code.

Prerequisites

Claude Code; authorized testing scope or bug bounty program; Docker recommended.

Use cases

securitypentestingbug-bountyagent-skills

Works with

Trust score — 6/10

/10 trustMetrics-checked

Scored 6/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 60; Trending score 60/100 from researcher assessment.

★ 563 GitHub stars

Sources

Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.