Make what your AI did explainable.
Any AI system logs its results as structured confidence, decisions, and actions_taken. What it did, why, and how sure it was — a searchable, filterable, exportable audit log.
Screens
What it does
- 01Append-only ledger
- No UPDATE or DELETE. Corrections are made by adding a new entry, never by rewriting history.
- 02Filter by confidence
- An overall score plus a per-criterion breakdown, so you can list the low-confidence entries first.
- 03Records the reasoning
- Not just what was chosen — why (decisions) and what was actually done (actions_taken).
- 04Ingest from any AI system
- comet-taskAI, ai-scheduler, your own scripts — anything that can POST over HTTP can log to it.
- 05Dashboard overview
- By-status and by-agent breakdowns, plus the low-confidence rate, at a glance.
- 06CSV / JSON export
- Export keeps your current search and filters — ready for audits and reports.
How it works
Launch and issue a key
Click “Issue key” in Settings. Copy the Ingestion URL and API key shown.
Configure your AI system
Give the URL and key to whatever should POST results — comet-taskAI, ai-scheduler, a custom script.
Watch the ledger and dashboard
Entries appear in real time as they arrive. Start with the low-confidence ones.
Good to know
- No UPDATE/DELETE API — a recorded entry can't be removed or altered.
- id and timestamp are server-assigned — client-supplied values are ignored.
- API keys are shown once and stored hashed, never in plaintext.

