Tools provided
Tools the agent exposes — your AI client calls them automatically when it needs them.
- run_full — Orchestrated path: synthesize your coded interviews end-to-end — theme codes with per-participant frequency, a per-research-question evidence table of timestamped attributed quotes, a separate counter-evidence section, and a narrated research report.
- analyze — Run pure deterministic analysis over the provided corpus with no model call and return scored findings.
- aggregate_themes — Aggregate theme codes with per-participant frequency from the provided corpus, most-mentioned first — the theming step on its own.
- build_evidence_table — Build the per-research-question evidence table: timestamped, attributed quotes split into supporting and contradicting.
- list_counter_evidence — Return only the counter-evidence section — every research question that has both supporting and contradicting evidence. The confirmation-bias guardrail as its own answer.
- synthesize_report — Run the full deterministic synthesis — themes, evidence table, counter-evidence, KPIs and findings — and return it with no language-model narration at all.
- list_capabilities — List the agent's static capabilities: the available tools and the analyzer's thresholds. 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 coded interviews and the research questions. The entry point for running the agent on your own corpus.
- run_form — Internal: invoked by the guided form when the user submits. Not for direct use.
