Part B · The screens
B.18Log
Every AI decision, with its inputs, the agent, the tool, the rule result, the model's tokens and the cost, and the AI spend this month against its budget.
- No.
- B.18
- Dashboard address
- /log
What you can do here
Each core task, and how many clicks it takes from this screen.
| Task | Clicks | How |
|---|---|---|
| See the AI spend this month | 1 | Log: spend so far, the monthly budget, the share used and the spend per day. |
| Filter the decisions | 2 | Log → a decision type (or pick an agent, a day or an account). |
| Open one decision | 2 | Log → the row: what went in, the result, the reasons and what could make it wrong. |
How it works behind the scenes
- Every action an agent takes is decided by the rule engine and written to the decision log first: allowed, held for a person, or refused, with the reasons.
- Model calls add their tokens and cost; the AI gateway stops model calls when the monthly budget is used up, and the AI calls kill switch stops them at once.
- The decision log is append-only: the database gives no role the right to change or delete an entry.
Rules and switches that apply here
- The monthly AI budget is a setting ($50 in the prototype); spending past it is refused by the gateway, not just reported.
The tables behind it
PK primary keyFK reference the database enforces (solid line)ID reference kept by id (dashed line)
| Table | What it holds | Columns |
|---|---|---|
| ctrl.decision_log | Every decision: the action, the rule result, the reasons, tokens and cost. | 28 |
| ops.llm_calls | Every model call behind those decisions. | 37 |
How the tables connect (7)
| Column | Points to | Kept by |
|---|---|---|
| ctrl.decision_log.agent_version_id | agent.agent_versions | The application (by id) |
| ctrl.decision_log.run_id | agent.runs | The application (by id) |
| ctrl.decision_log.conversation_id | conv.conversations | The application (by id) |
| ops.llm_calls.agent_version_id | agent.agent_versions | The application (by id) |
| ops.llm_calls.run_id | agent.runs | The application (by id) |
| ops.llm_calls.conversation_id | conv.conversations | The application (by id) |
| ops.llm_calls.decision_id | ctrl.decision_log | The application (by id) |
The screen
localhost:13301/log
Log: the AI spend this month, and every decision (live).
- 1Spend so far against the budget, and per day.
- 2One decision: when, what, which agent and tool, the rule result, tokens and cost.
Dashboard · Govern → Log, AI spend. Captured from the running POC on September 25, 2026; synthetic data.
See it live
Opens the screen in the running POC, in a new tab.
| Open | Where | What to try |
|---|---|---|
| Log | DMA AI dashboard | Spend this month, then every decision underneath. |