rokha-ai/rokha
Search the Rokha registry of agent skills and MCP servers, then run them for real — no install, a trace on every run.
A ready-made three-step Rokha Rig pattern — fetch → analyze → report. It takes ONE registry listing as its input: step 1 looks that listing up and fetches its real document, step 2 inspects the document for security and trust-boundary issues, step 3 writes a plain-language brief ending in an adopt / try-with-care / avoid verdict. Use when the user wants to vet a tool before adopting it, or wants a worked example of a fetch → analyze → report workflow. Instantiable by any agent over Rokha's public API, or one click in the Rokha editor.
A ready-made three-step Rokha Rig pattern — pick a listing from the Rokha Registry (a skill, harness, or rig), dissect its real-world use cases and best pairings, and write an explorer's brief. Use when the user wants to "check out a tool", understand what something in the registry is for, or wants a worked example of a lookup → analyze → brief workflow. Instantiable by any agent over Rokha's public API, or one click in the Rokha editor.
A ready-made four-step Rokha Rig pattern chaining the Explore and Audit patterns into one flow — pick a Rokha Registry listing, dissect its use cases, vet its safety, and land on an adopt / try-with-care / avoid verdict. Use when the user wants a full evaluation of a tool before adopting it, or a worked example of a multi-step lookup → analyze → vet → verdict workflow. Instantiable by any agent over Rokha's public API, or one click in the Rokha editor.
A ready-made two-step Rokha Rig pattern — step 1 writes a piece (any writing skill from the live registry), step 2 critiques it against the actual output. Use when the user wants to "write and review", "draft then critique", or wants a worked example of chaining two skills into one workflow. Instantiable by any agent over Rokha's public API, or one click in the Rokha editor.
Security and compliance audit for MCP tools, agents, and skills found in the wild. Run before installing or invoking any unfamiliar tool. Produces a clear "safe to use / here's how" or "found these vulnerabilities" verdict with the schema, endpoint, install command, and trust signals the agent or user needs to make a decision. Use when the user mentions auditing, vetting, security-checking, or "is this safe" about an MCP tool, ClawHub skill, or Smithery server.
Search the Rokha Registry's 205k+ published agent skills and install any of them into your own skill library — no account, no API key. Use when the user asks to find an agent skill ("is there a skill for X?"), browse what skills exist for a task, or download/install a skill from the Rokha registry (or clawhub) into Claude's skills. Works over Rokha's public MCP endpoint with plain HTTP calls.