Tools provided
Tools the agent exposes — your AI client calls them automatically when it needs them.
- run_full — Orchestrated path: age the provided ledger, score per-customer risk, draft the tone-graduated reminders, propose escalations and forecast collection.
- analyze — Run pure deterministic aging and risk analysis over the provided ledger with no model call, returning the full aging report.
- age_ledger — Age the ledger into the standard buckets — current, 1-30, 31-60, 61-90 and 90 plus — with the bucket totals and the outstanding and overdue figures.
- score_risk — Score per-customer collection risk from worst-bucket exposure, overdue ratio and broken payment promises.
- draft_reminders — Draft one tone-graduated payment reminder per overdue customer — friendly, firm, final or escalation, matched to how overdue the account is. Drafts only; sending requires human approval.
- propose_escalations — Return escalation recommendations for the worst accounts — those over 90 days or with repeated broken promises.
- forecast_collection — Return the probability-weighted 30-day collection forecast, computed from the aged ledger.
- list_capabilities — List the agent's static capabilities: the available tools, the aging buckets and the collection-probability model. 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 collections goal and co-design the exact input through clarifying questions before running.
- open_form — Opens the guided form to collect your open-invoice ledger. The entry point for running the agent on your own receivables.
- run_form — Internal: invoked by the guided form when the user submits. Not for direct use.