SnailSploit/Claude-Red

What it is

claude-red is a library of offensive security skills for the Claude Skills system. Kai Aizen (SnailSploit) maintains it under the MIT license. Each skill is a SKILL.md file under Skills/<category>/, for example Skills/web/offensive-sqli/SKILL.md. The README lists these use cases: authorized red team engagements, bug bounty triage, security research, CTF preparation and operator training. The library is still being built toward about 130 skills, and the page points to CHANGELOG.md for release history. It does not give the current skill count.

What to do

Full install, which the page recommends:

git clone https://github.com/SnailSploit/claude-red ~/.claude/skills/claude-red

Claude then loads matching skills on conversational triggers. The page says skills load on demand, so unused ones cost no context.

To install only some categories, use a sparse checkout:

git clone --filter=blob:none --sparse https://github.com/SnailSploit/claude-red
cd claude-red && git sparse-checkout set Skills/web Skills/active-directory

The install script has three forms:

./install.sh                           # interactive
./install.sh --target ~/.claude/skills # explicit target
./install.sh --category web            # single category

Without the skills system, the page shows two manual routes. One pipes a skill into the CLI as a system prompt:

cat Skills/web/offensive-sqli/SKILL.md | claude --system-file -
cat Skills/active-directory/**/SKILL.md | claude --system-file -

The other pastes a SKILL.md into a Project's system prompt or prepends it to a conversation.

Caveats

• The page gives no sample skill content, so the depth and accuracy of the techniques are unverified from the README alone.
• Contributions follow CONTRIBUTING.md, which sets a skill template and frontmatter standard. It prefers focused single-surface skills over monolithic overviews.
• Many skills build on Sahar Shlichov's offensive checklist collection.
• The README frames use as authorized engagements, but it does not describe any scope controls or guardrails in the skills.