ODF docs
GitHub ↗
Documentation menu

ODF Fast-Lane Baseline — 2026-09-16

2 min read

#Evidence status

  • Observed command: node scripts/odf-toolkit.js metrics --days 14 --json on 2026-09-16 UTC.
  • Purpose: FL-01 task-call evidence and the FL-02 cohort/rollback definition for ODF2 readiness.
  • Claim boundary: This report measures phase/task duration_ms. It does not measure /odf-new or /odf-fix entry-to-final-gate latency and makes no performance claim.
  • Odoo runtime: Odoo version, project, and a real Odoo runtime are not present in the sanitized telemetry.

#Cohort and window

  • Window: 14-day window selected by the command.
  • Raw telemetry records: 418. This is the raw-record count, not the baseline denominator.
  • Baseline completed task-call samples: 133. Lifecycle markers, spans, and joins are excluded from this denominator.

#Overall task-call baseline

Sample p50 p95 Average
133 123,671 ms 600,847.6 ms 250,187.5263 ms

#Outcomes

Outcome Count
ok 29
blocked 67
error 12
timeout 25

#Phase/task duration

These are phase/task durations, not /odf-new or /odf-fix entry-to-final-gate durations.

Phase p50 p95
ASSESS 361,301 ms 750,361 ms
DESIGN 455,890 ms 634,302.6 ms
EXPLORE 13,819 ms 13,819 ms
FIX 13,063 ms 1,446,798.95 ms
IMPLEMENT 54,034 ms 600,533.8 ms
PROPOSE 101,114.5 ms 178,640.2 ms
QA-PLAN 241,476 ms 600,373.1 ms

FIX is a phase label, not proof of an end-to-end /odf-fix run.

#Dimension coverage

  • Work type: 70/133 (52.63%) reported: small-change 12, feature 52, and bugfix 6. The remaining 63 samples are unknown.
  • Model field: 133/133 present, but model_available is 0/133. Provider/model identity is unavailable.
  • Validation evidence: 0/133.
  • Receipts: 0/133.
  • Escalation: 0/133.

Coverage gaps are: work_type, model_identity, validation_evidence, receipt, escalation, and entry_to_final_gate.

#End-to-end measurement gap

entry_to_final_gate is unavailable: sample_count is 0 and p50/p95 are N/A. The telemetry therefore cannot establish complete /odf-new or /odf-fix entry-to-final-verified-gate cohorts. This report makes no end-to-end or performance claim.

#FL-02 proposed cohort and rollback definition

Comparable cohorts must contain complete entry_started → verified_completed flows with matching work_type, Odoo version, workspace, and source-authority dimensions when available. Lifecycle markers, spans, and joins are excluded from the denominator. No comparable cohort exists in this evidence window.

The following hold/rollback signals are proposed and require explicit maintainer approval before any canary:

  • 10% matched-cohort p95 regression.

  • Any protected or architectural risk incorrectly downgraded.
  • Any candidate, evidence, or receipt binding failure.
  • Missing final-gate evidence.
  • Telemetry coverage below 95%.
  • Any unexplained increase in errors, timeouts, or receipts.

No canary, rollout, or ODF2 execution claim is made.