Primary Drift Patterns
- Worker and main-thread identity mismatch under extended sessions.
- Connection narrative instability after proxy pool rotation.
- Governance drift when seat count increased without role policy hardening.
April 2026 Reliability Baseline
This report establishes the first monthly baseline across 22 platforms using methodology v1.2. It is designed for shortlist refinement, pilot planning, and risk-aware procurement review.
Protocol Context
How This Benchmark Was Run
| Dimension | Value |
|---|---|
| Methodology version | v1.2 reliability-weighted protocol |
| Critical run requirement | 3 repeated clean runs per critical workflow |
| Core scoring weights | Profile integrity 38%, API reliability 26%, operational cost 21%, team governance 15% |
| Publication blockers | Critical drift, unresolved connection leaks, missing no-buy criteria |
Headline Findings
April 2026 Summary
8Platforms with Level A evidence in baseline flows
9Platforms with Level B evidence and explicit caveats
5Platforms with no-buy blockers in at least one critical flow
14Profiles showing measurable governance drift risk under team scaling
Top-Line Matrix
Reliability Bands by Evidence Level
| Band | Score range | Evidence posture | Procurement interpretation |
|---|---|---|---|
| Band A | 8.5 to 10.0 | Level A with stable repeated runs | Eligible for controlled procurement after local validation |
| Band B | 7.0 to 8.4 | Level B with caveat-bound reliability | Pilot-first, avoid annual commitment until caveats are closed |
| Band C | 0.0 to 6.9 | Level C or unresolved blockers | No-buy for critical workflows at current baseline state |
This table is directional. Always validate with your own stack, proxy profile, and team operating model.
Decision Impact
- Budget-first stacks require stronger rollback discipline before annual contracts.
- API-heavy teams should prioritize lifecycle reliability over entry pricing.
- Promo claims should be treated as secondary until evidence reaches stable Level A/B.
Action Path
What to Do After Reading This Report
Step 1: map your shortlisted candidates against this month baseline bands.
Step 2: run Fingerprint Readiness Score to pre-check pilot risk.
Step 3: execute ops SOP gates and lock no-buy criteria.
Step 4: generate a reusable evidence pack for approval traceability.
Step 5: only then move through compare and promo routes for procurement action.
Evidence Governance