Skills
97,725 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
Show a plan's status as one readable offline HTML page — what the project is, how far it got, what waits on the user, what changed since approval, and what the system judged for them — built from SPECs, tickets, change proposals, decision logs, and review verdicts. Refresh `.scratch/<feature>/plan.html` when it exists and a stage finished, work got blocked, or an acceptance result arrived. Use when HTML is requested or plan relationships need a visual overview; plain plan wording alone is not a trigger
Use when the user asks for a retrospective on an agent session ("/retro", "how did this session go", "what could this session do better"), for the current session or a given session ID or transcript path. Reports time per phase, tool-call errors, repeated slow test runs, and skill inefficiencies. Works for Claude Code and pi.
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.
