betool

Documentation

Execution History

Browse, filter and inspect pipeline executions node by node, then replay them as a dry run or for real. History is kept with no automatic purge.

Every pipeline trigger creates an execution recorded in the database. You can inspect it to the second, see what each node produced, and replay it.

Accessing the history

From Projects (side menu), open the project concerned: each conversation unrolls its exchanges and their executions.

The table is filterable by:

  • Time period
  • Status (success, error, in progress, abandoned)
  • Input channel
  • Cost (in credits)

Anatomy of an execution

Click a row to open the execution drawer. You see:

  • The complete input payload
  • The traversed graph — each node visited, with its state (skipped, ok, error)
  • For each node:
    • Its inputs at the time of execution
    • Its outputs produced
    • The execution duration
    • The cost (for metered nodes: agent, voice, etc.)
  • The tools called by each agent, with their arguments and return values
  • The complete LLM log (prompts sent, responses received)

Replay

From the drawer, replay the exchange on the pipeline's published version, in two modes:

  • Simulation (dry-run) — the execution runs with no external side effects. Useful for reproducing a bug or validating a fix without sending anything.
  • Real — the full execution, side effects included.

The replay is tracked as a new execution. The original history remains intact.

Advanced filtering

The table supports native filters by single value or date range. You can also export the filtered executions as Excel / PDF for sharing outside the platform.

Retention

Execution history (conversations, exchanges, node traces) is retained and remains accessible — no automatic purge is applied to it.

Two internal technical logs, on the other hand, are automatically purged after 30 days: the queue of inbound events already processed (failed ones are kept), and the idempotency register of confirmed outbound sends. They serve exactly-once delivery, not consultation.

For targeted erasure (GDPR), see GDPR Audit.

Audit use case

If an auditor asks you to "show the exchanges where the agent refused a request":

  1. Filter by project / pipeline and by period.
  2. Open the executions concerned: every tool call appears there with its arguments and return value.
  3. Export the selection as Excel / PDF.

Everything is traceable, nothing is fabricated.