{"name":"ago","title":"AOO — Agent Operability Optimizer","version":"2.0.0","description":"Measure whether AI agents can operate a website.","transport":{"type":"streamable-http","url":"https://daas-aoo.listeningmind.ai/api/mcp"},"tools":[{"name":"get_score","description":"Look up a domain's stored agent-operability score without running anything. Try this before scanning — a scan drives a real agent for minutes. Returns null if the domain has never been scanned."},{"name":"scan_domain","description":"Run a full agent-operability audit: several real agents attempt real tasks on the site and every request is instrumented. Returns immediately with an audit_id — poll get_audit until status is \"done\" (typically 1-3 minutes)."},{"name":"get_audit","description":"Fetch a full audit report: category scores, every goal with its evidence, findings worst-first, recommendations, and per-task results. Returns status \"running\" while the agents are still working."},{"name":"run_trace","description":"Send one agent at a site on a single task and record its full trajectory. Use this to re-verify one thing after a fix, instead of paying for a whole audit. Returns a run_id — poll get_trace."},{"name":"get_trace","description":"Fetch one recorded trajectory. Every step carries an attribution saying HOW the agent learned that URL — link, sitemap, llms_txt, robots_txt, earlier_response, web_search, model_recall. This is the raw evidence every score sits on."},{"name":"list_goals","description":"The catalog behind the score: every goal, its category, and how it is measured. Goal ids are stable — gate CI on explicit ids rather than on category totals."},{"name":"get_leaderboard","description":"Domains AOO has scanned, ranked by their latest audit. This ranks scanned sites, not the web — absence from the board means \"never scanned\", not \"worse\"."},{"name":"list_scanned_domains","description":"Every domain scanned so far with its latest score and scan count."}],"documentation":"https://daas-aoo.listeningmind.ai/docs"}