Skills

Each skill installs into ~/.claude/skills/ (or the equivalent for your agent runtime) and is invocable by name or by asking naturally.

• Solana Risk & ComplianceOpen Source · MIT

  Turns Claude Code (or any compatible coding agent) into a three-person risk team — token/wallet screening, securities & AML triage, and a six-step forensic investigation method. Boots keyless on public infra; deepens automatically if a RugBurn API key is present.

The risk & compliance skill is self-contained — no API key required to start. Cross-platform installers for macOS, Linux, and Windows:

git clone https://github.com/storm-beyndtech/solana-risk-compliance-skillcd solana-risk-compliance-skill./install.sh          # macOS / Linux, interactive (-y for non-interactive)pwsh ./install.ps1     # Windows (PowerShell)

Restart your agent (or /reload) so it picks up the skill. If a RUGBURN_API_URL and RUGBURN_API_KEY are set, the agents use them for calibrated behavioral scores and deployer reputation; without them, every check still runs from first principles against public RPC.

We build integration skills with design partners as those integrations ship — this page is where they will appear. Nothing to show here yet beyond the risk & compliance skill above.

Skills return screening evidence and calibrated findings for an agent to reason over. They are not financial, legal, or investment advice, and regulatory output is not a substitute for licensed counsel. Every skill in this list defers final decisions to a human — see each repository's own human-in-the-loop rules.