I3299 — Majordomus's own overhead in an adopting repository is measured per operation against declared budgets

○READY wave 0 · p3 · implementation profile · parallel safe

Part of every-sessions-usage-is-observed — What every AI session consumed is observed from the provider, attributed to the work it served, and never estimated into a fact.

Objective

Declare budgets for CLI start, warm discovery, idle server CPU and resident memory, incremental derive, and the bootstrap tokens a session pays, measure them in an adopting fixture repository with the existing bench, and fail when one is exceeded.

Why

The owner asked for an honest answer on whether overhead stays acceptable; the proposed numbers (2026-10-16) are targets, not measurements, and nothing measures them in a foreign repository today. It serves the intent's own-overhead-is-measured, the proof of the statement's 'costs'; it is the adopting-repository delta beside I2485, ADR 0169 and project.performance-evidence, which measure this repository only, and it stays bound to this intent because milestone development-runtime belongs to no intent. It extends milestone development-runtime. Sources: chatgpt-majordomus:c278#2, chatgpt-majordomus:c278#3.

Current state

rule project.performance-evidence and bench --check hold UI latency budgets (I2489 done) for this repository only.

Desired state

An adoption overhead suite in benchmark policy with recorded measurements.

Scope

• .ai/repo/benchmarks/rust/policy.yaml
• .ai/repo/benchmarks/rust/README.md
• .ai/repo/benchmarks/README.md
• apps/majordomus-cli/src/bench/system.rs
• apps/majordomus-cli/src/bench/runner.rs
• apps/majordomus-cli/src/bench/mod.rs
• apps/majordomus-cli/src/capability/benchmark.rs
• apps/majordomus-cli/src/commands/bench.rs
• apps/majordomus-cli/tests/bench.rs
• docs/generated/benchmarks.json

Dependencies

None. This issue is a root of the graph.

What waits on this

• I3364✕BLOCKED The adoption-overhead bench takes a target repository, and Majordomus's overhead and footprint are measured in the adopted foreign repository

Acceptance criteria

• Each budget has a benchmark that fails when exceeded
• Idle CPU is measured on an inactive fixture

Validation

• majordomus bench --check
• majordomus plan validate

Evidence required

• bench_adoption_overhead

Evidence

None recorded. Every token above needs a command or an artifact behind it before this issue can be completed; narrative is refused.

Timeline

started

—

verified

—

completed

—

Those three fields, the evidence above and the state of the dependencies are all the status is made of. There is no status field to disagree with them.

Canonical record: .ai/repo/project/issues/I3299.yaml. Read it back with majordomus plan show I3299.