I3388 — The mesh suite is registered before its first run, and the T4 protocol names its registration digest

✕BLOKOVÁNO vlna 7 · p1 · profil implementace · bezpečné paralelně

Part of the-mesh-is-measured-against-one-agent — Coordinated workers on the mesh are compared with a single agent on the same batch tasks, at quality held by a registered non-inferiority guard.

Cíl

Write .ai/repo/benchmarks/economics/suites/mesh.yaml as a registered suite: the batches, control single agent with the pinned release, treatment coordinated workers with the registered worker count, the same model, harness and per-session budget, wall time to accepted integrated tree as primary outcome, the accepted-share guard and margin, the MDE, the run-order seed and the stopping rule with the owner's budget decision. Register the T4 paired protocol in the proof plan.

Proč

T4 of the owner's report of 2026-10-19; the quality guard and worker count fixed before any run keep the comparison from being tuned to its result.

Aktuální stav

git ls-files .ai/repo/benchmarks/economics/suites lists context.yaml and pilot.yaml.

Cílový stav

economics summary lists suite mesh with every pair missing; T4 reads not_run with reason no counted record.

Rozsah

• .ai/repo/benchmarks/economics/suites
• .ai/repo/benchmarks
• docs/generated/economics.json
• docs/generated/economics.md
• docs/generated/economics.yaml
• site/data/generated/economics.json

Mimo rozsah

• Running the suite (I3389)

Závislosti

• I3385✕BLOCKED A registered suite's primary verdict can be held to a non-inferiority guard, so a speed-up counts only when accepted share is not worse by more than a registered margin
• I3386✕BLOCKED The economics runner runs one job as several concurrent workers in worktrees of one workspace sharing the treatment's server, and records the job's wall time to an integrated, accepted tree
• I3387✕BLOCKED Batch tasks compose existing corpus tasks into sets of issues several workers can share, with their dependencies and the union of their hidden tests
• I3352✕BLOCKED The six proof tests are registered, each with the claim it settles and its instrument, and none carries a protocol yet

Co na tom čeká

• I3389✕BLOCKED The mesh suite is run in both arms and its guarded verdict is published with its losses

Kritéria přijetí

• The registration loads with its guard and margin
• The T4 protocol validates with the registration digest
• The budget ceiling cites a recorded owner decision

Validace

• majordomus economics summary
• majordomus economics proof
• majordomus generate --check
• cargo test --manifest-path apps/majordomus-cli/Cargo.toml --test economics

Požadované důkazy

• mesh_registration_committed

Důkazy

Nic nezaznamenáno. Za každým tokenem výše musí stát příkaz nebo artefakt, než lze tento úkol dokončit; vyprávění se odmítá.

Časová osa

zahájeno

—

ověřeno

—

dokončeno

—

Tato tři pole, důkazy výše a stav závislostí jsou vše, z čeho se stav skládá. Neexistuje žádné pole stavu, které by s nimi mohlo být v rozporu.

Canonical record: .ai/repo/project/issues/I3388.yaml. Read it back with majordomus plan show I3388.