refine_results

com.agentery/agentery · Verify this server

Refine an earlier search_providers or research_capability answer with a correction, and get a better answer back in the same call. Pass the request_ref returned with that answer and the correction in plain words: what was missing or wrong (e.g. 'I meant apps that reduce business costs', 'must integrate with Slack', 'budget is $20/month', 'data must stay in the EU'). Agentery re-reads the request with the correction (changed constraints change, the others are kept), runs the search again and returns the new answer plus a short explanation of what changed. Optional: source = 'human' (the user's own words), 'agent' (your own assessment) or 'follow_up' (an ordinary narrowing, not a complaint). A message with nothing actionable (e.g. 'thanks') is just acknowledged. Agentery only sees the request and correction you send, never your wider conversation. Task answers (search_providers, compare_providers, research_capability, price_benchmark, market_report) carry ONE request_ref plus a 'conversation' block: the facts remembered for the task, any optional clarification questions (at most two, one round) and the exact refine_results call to continue. Answer with plain words, 'unknown' or 'skip'; source is one of human, agent, follow_up. Pass the same request_ref to another task tool to continue the task there. References expire 24 h after the last turn.