Product / Evaluations

Test agent behavior against policy.

Decisions are deterministic, so agent behavior is testable. Assert the answers agents should get against published policy, review every question they asked, and gate SQL changes with query validation.

Will an agent get the right answer? Check before it ships.

Red-team cases state the decision an agent should get — allow, deny, or conditional — and run against the live MCP tools. When policy changes, the same checks re-run and fail loudly if behavior shifts.

Red-team case · export PII to CRM
deny
Matches expected decision
Red-team case · revenue aggregation
allow
Matches expected decision
Release gate · changed SQL
pass
No restricted tables for intent
passing
All expected decisions match → Suite passes

Evaluations

Red-team

Expected decisions, asserted

Write cases that state what an agent should be told — this use allowed, that export denied, this one conditional. Run them against the live MCP tools and assert the typed answer state, deterministically, in CI.

Request log

Review what agents actually asked

The per-token request log shows every question agents asked and the typed answer state each got — 90 days, allow-listed fields, query text never recorded.

CI gate

Query validation as a release gate

Run SQL and data changes through validate_query_context before merge. The check is fail-closed: if a changed query touches a restricted table for its intent, the gate flags it before release.

In development

Not yet shipped

Today, decisions are re-derived from current serving state and validation answers are not persisted. The next evaluation layers build on a durable decision ledger:

Decision-ledger evaluations

Score recorded decisions against policy after the fact.

Drift trends

See decision patterns shift across runs before they compound.

Policy replay

Re-run past decisions against a new policy version to see what would change.