a-registered-corpus-judges-majordomus — The registered corpus judges the same agent with and without Majordomus, by an exact task-level test on accepted results, and the verdict is published with its losses

○PLANNED 0 of 4 issues done

ready

2

blocked

2

active

0

verify

0

done

0

cancelled

0

Problem

Whether Majordomus delivers more accepted work than the same agent without it has never been measured. Under methodology version 1 a pair with a failed side is invalid (methodology.yaml:44), so acceptance cannot be compared in pairs; stats.rs:260 offers only a clustered interval of a median; no live run has ever been committed. I3133, which declares the acceptance pairing, is BLOCKED through I0401, whose milestone cost-per-accepted-outcome depends on execution-telemetry (BLOCKED), so T2 cannot finish inside the owner's 30-day window as the plan stands.

Outcome

The blocking chain is cut or re-scoped by recorded decisions; the calculator decides the primary outcome by an exact task-level test and reports secondary outcomes without a verdict word; the corpus is run in both arms and the verdict, losses included, is published on the T2 row.

Current state

site/data/generated/plan.json derives I3133 BLOCKED (depends_on I0401), I0401 BLOCKED (milestone cost-per-accepted-outcome depends on execution-telemetry), execution-telemetry BLOCKED. git ls-files .ai/repo/benchmarks/economics/runs lists only runs/context (9 records). EconomicsSession.duration_ms (model.rs:687) is recorded and read by no metric. claims.rs:41-44 VOCABULARY holds no acceptance or wall-time word.

Desired state

economics summary reports the registered suite's primary verdict, its secondary differences with intervals, attempted, valid and missing pairs; runs/corpus holds every attempted run; the T2 row reads a computed status.

Scope

• apps/majordomus-cli/src/economics/stats.rs
• apps/majordomus-cli/src/economics/mod.rs
• apps/majordomus-cli/src/economics/model.rs
• apps/majordomus-cli/src/economics/report.rs
• apps/majordomus-cli/src/economics/claims.rs
• apps/majordomus-cli/tests/economics.rs
• share/schemas/generated/economics-summary.schema.json
• site/templates/economics.html
• .ai/repo/benchmarks/economics/runs
• .ai/repo/project/issues
• .ai/repo/project/issues/I0401.yaml
• .ai/repo/project/milestones/delivery-flow-is-measured.yaml
• docs/generated/economics.json
• docs/generated/economics.md
• docs/generated/economics.yaml
• site/data/generated/economics.json
• docs/ECONOMICS.md
• site/content/docs/economics.md

Out of scope

• The definition of accepted (I0401) and the acceptance pairing (I3133)
• Arms per mechanism (I3298)
• Human interventions and CI reruns on real work (T3)
• Re-running or excluding a recorded run to change a verdict

Acceptance criteria

• Each blocked dependency of the program is either cut by a recorded owner decision or recorded as refused with the criteria it leaves outside the window
• The primary verdict is computed by the registered exact test and is inconclusive when underpowered
• Every declared pair is attempted and every missing or invalid pair is published with its reason

Validation

• majordomus plan validate

Risks

• The owner may refuse to cut I3133's dependency; then T2 waits on execution-telemetry
• Live runs spend provider usage and are account-bound; a weekly limit can leave pairs missing, which is published as missing
• With 20 tasks the acceptance difference may well be inconclusive; that is a valid published result
• I3133 and I3298 scope the whole economics directory; the landing seat orders them against I3366 and I3371

Issues

Each one is an execution contract with its own acceptance criteria, validation command and required evidence. The wave column is the layer of the dependency graph it sits in.

The graph

flowchart LR
    I3366["I3366<br/>The economics calculator decides a binary paired outcome by an exact task-level test, intervals a continuous paired difference by task-clustered bootstrap, and turns either into one verdict"]:::ready
    I3371["I3371<br/>Economics summary decides a registered suite's primary outcome by its registered test, reports secondary differences without a verdict word, and refuses typed figures about them"]:::blocked
    I3372["I3372<br/>The registered corpus is run in both arms, its records are committed, and its verdict is published with its losses and its origin"]:::blocked
    I3373["I3373<br/>The program's blocked dependencies are cut or re-scoped by recorded owner decisions, or recorded as refused with the criteria they leave outside the window"]:::ready
    I3366 --> I3371
    I3371 --> I3372
    classDef done stroke:#16a34a,fill:#dcfce7,color:#14532d,stroke-width:2px
    classDef active stroke:#2563eb,fill:#eff6ff,color:#1e3a8a,stroke-width:2px
    classDef verify stroke:#7c3aed,fill:#f5f3ff,color:#4c1d95,stroke-width:2px
    classDef ready stroke:#0891b2,fill:#ecfeff,color:#164e63,stroke-width:2px
    classDef planned stroke:#0891b2,fill:#ecfeff,color:#164e63,stroke-width:2px
    classDef blocked stroke:#b45309,fill:#fffbeb,color:#78350f,stroke-width:2px
    classDef cancelled stroke:#6b7280,fill:#f9fafb,color:#374151,stroke-width:2px
    classDef superseded stroke:#6b7280,fill:#f9fafb,color:#374151,stroke-width:2px

Canonical record: .ai/repo/project/milestones/a-registered-corpus-judges-majordomus.yaml. Read it back with majordomus plan show a-registered-corpus-judges-majordomus.