Surface Notice|Validation Lab is the public evidence adjudication surface for MPLP. It publishes evidence-based verdict views, not certification, compliance marks, or execution hosting.
Snapshotrc-20260131161432
Inventory Date2026-01-31
operational(inferred)

Multi-Agent Lifecycle Protocol

Validation Lab

Non-Normative Evidence Adjudication Lab

What you’re seeing This lab is currently operational. It is a bounded evidence adjudication surface for MPLP, not a certification program or protocol-definition authority.

↓ Download an evidence pack → recheck locally → compare the verdict hash.

We recheck evidence integrity and deterministic adjudication outputs — not runtime performance, not certification.

Start Here

Browse Runsets

Choose a public evidence series

Release History

Release Records

Review sealed public records

Learn

How Verdicts Work

Ruleset logic & clauses

Protocol Ref

v1.0.0 Frozen

Docs Entry Points ↗

Integrity

System Sealed7/7 Gates Passing

View Records →

Series

v0.17.0 Sealed

View History →

Integrity Status: SEALED
Cross-Verification View →
Validation HubCross-Verification ReportRulesets
⚠

Reviewability ≠ Reproducibility. The Lab rechecks evidence, not execution.

Quick Start

Start with: gf-01-langchain-passRECOMMENDED

# Verify locally (read-only)

npm run vlab:recheck-hash gf-01-langchain-pass

Start with Runsets →All runsHow to Recheck Evidence ↗

The Audit Loop

Download an evidence pack → recheck locally → compare the verdict hash.

  1. 1Browse Runsets

    Pick a run

  2. 2View Evidence

    Inspect pack

  3. 3Download Pack

    Get JSON/Zip

  4. 4Recheck Evidence ↗

    Run recheck

  5. ✓Compare Hash

    Match Seal

Discovery Resources

Validation Hub

Gate feeds, integrity checks, and bounded Lab navigation.

Cross-Verification

Source-bound comparison snapshot over a historical cross-verification report.

Runs

Curated evidence packs including PASS, FAIL, and NOT_ADMISSIBLE.

Rulesets

Versioned adjudication lines and source-bound decision logic.