MCP endpoint
https://compass.ariobarin.com/mcp
- blast-radius Find what a change could break somewhere else before it ships, beyond the diff, and prove the one fact it's safe because of by running real code instead of writing it up. Use for 'blast radius of X', 'what could this break', or reviewing a small diff you don't trust.
- compass Change Compass source or refresh the live Compass installation. Use when the user wants a behavior, skill, agent, prompt, configuration, manifest, installer, or installation update handled through ariobarin/compass.
- design-prototype Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.
- diagnosing-bugs Observed software failure or performance regression requiring diagnosis. Use only when the user reports a concrete broken, failing, throwing, crashing, flaky, incorrect, or unexpectedly slow symptom, or explicitly asks to debug or diagnose one. Reproduce that symptom with a tight red-capable command before forming a theory.
- grilling Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.
- ground-in-sources Extract hard-won doctrine from people who use, test, and suffer the thing. Use before writing durable prompts, skills, agents, plans, policies, or technical direction that should inherit field experience instead of vendor-filtered or invented guidance.
- handoff Compact the current conversation into a handoff document for another agent to pick up.
- micro-experiment Test an uncertain product or technical idea with disposable code before integrating it. Use for proofs of concept, spikes, prototypes, novel pipelines, unfamiliar APIs, or any idea whose shape, feasibility, behavior, or value is still unknown.
- monitor Supervise long-running processes, downloads, updates, benchmarks, scripts, agents, and external waits without burning model turns on repetitive polling. Use when work outlives a quick tool call and needs continued health checks, recovery, or intervention before terminal evidence arrives.
- no-comments Spawn Comment Sicko, fix accepted findings, and offer encodings for claimed constraints.
- test-for-risk Choose evidence proportional to regression risk and maintenance cost. Use when deciding whether a change needs tests, selecting test scope or level, repairing bloated or brittle tests, designing reusable test infrastructure, or deciding what belongs in recurring CI.
- to-tickets Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker (edges as text in one file per ticket locally, or native blocking links on a real tracker).
- unslop Remove canned AI-writing patterns from explanations, documentation, PR text, and public prose. Use when a draft is inflated, formulaic, vague, or repetitive, or when the user asks to unslop writing.
- write-a-skill Create, revise, prune, or retire an agent skill. Use when a recurring capability or behavior needs selective triggering, high-signal guidance, reusable resources, composition with other Compass skills, or proof on realistic work.