Tools provided
Tools the agent exposes — your AI client calls them automatically when it needs them.
- run_full — Orchestrated path: score sentiment, extract topics, rank the posts worth a reply and compute the crisis risk over the mentions you provide, returning the daily digest and any urgent alert.
- analyze — Run pure deterministic analysis over the provided mentions with no model call and return the full digest.
- score_sentiment — Return the sentiment distribution alone — positive, neutral and negative counts with the negative ratio.
- extract_topics — Return the prominent topics alone, drawn from your keywords and the hashtags actually appearing, most-mentioned first.
- recommend_responses — Return the five posts most worth a human response, ranked by sentiment multiplied by reach, each with a suggested angle. Guidance only — nothing is ever sent.
- score_crisis — Return the 0 to 100 crisis risk score and its band — low, elevated, high or critical.
- summarize_volume — Return mention volume alone — the total, broken down by source and by whether each mention is about you, a competitor or the category.
- list_findings — Return the worst-first findings alone: crisis risk, negative sentiment share, high-reach negatives and volume.
- list_capabilities — List the agent's static capabilities: the available tools and the crisis scoring bands. Useful for discovery.
- plan_inputs — Plan the inputs for a tool: returns the clarifying questions, the JSON schema and a ready-to-edit example.
- discover_intent — Understand your goal and co-design the exact input through clarifying questions before running.
- open_form — Opens the guided form to collect the brand, competitors, keywords and the mentions to analyse. The entry point for running the agent on your own data.
- run_form — Internal: invoked by the guided form when the user submits. Not for direct use.
