Call external APIs, MCP servers, and CLI tools by writing TypeScript code that runs in a sandbox — without MCP. Use when the user wants the agent to use a tool/API/integration (e.g. "search Open Library", "read my Notion", "call this REST API") in an environment that has a shell but no MCP config and no settable…
Call the MCP servers this machine's coding agent is already authenticated to — from a shell script or from TypeScript, with no model in the loop. Use when a task needs data from an MCP server (Linear, Slack, Axiom, Fastmail, a self-hosted one), when several MCP calls should be composed into one script instead of a…
Develop and extend the Audius Code Mode MCP server — the Effect-TS server that exposes the Audius and Open Audio Protocol APIs to LLMs via search + execute tools. Use this skill whenever the user wants to add features, fix bugs, refactor, or extend this MCP server. Trigger on mentions of MCP server development, adding…
Guide for designing, implementing, and advising on MCP servers that use Code Mode — the pattern where an LLM writes and executes code to orchestrate API calls instead of calling individual tools one at a time. Use this skill whenever the user mentions Code Mode, MCP tool proliferation, context window bloat from MCP…
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Drive a Parselbox sandbox effectively. Use when a Parselbox executecode MCP tool is available, when writing code against the parselbox Python SDK, or when wiring parselbox into an MCP config. Covers discovery (sbx.), stateful execution, background tasks, bash/js/require interop, filesystem, and serve/@api patterns.