AI Development Methodology

E00 — Intake

Pillar (primary): P9 — Self-improvement velocity Status: active (2026-08-25) — standing; this epic does not close Phase: all phases Started: 2026-08-25 Target close: never Owner: maintainer

Frozen intent — Outcome and exit criteria approved by maintainer on 2026-08-25. Agents do not edit them; halt and renegotiate instead.

Outcome (jobs-to-be-done)

When work arrives that is real, small, and not worth a charter — a short-notice task, a changed requirement, a decision recorded in one pass and left unexecuted — the maintainer wants somewhere to file it in seconds under the full item discipline, so that not planning it does not also mean not gating it.

This is the dogfood run for “focused mode.” The idea is described in FUTURE.md; it is deliberately not published in methodology/ and will not be until this repo has used it on real work. Sixteen conventions shipped v1.25.0 → v1.31.0 and only seven have ever been exercised. This one gets used before it gets written.

What intake is, in one line

Same item, no charter. A BL-#### filed here carries the full eight-field table, frozen intent on its Done means:, the Status/Test hard rule, the six DoD gates, the lock protocol, and a Code Map at Effort M+ — identical to a cascade item. The only thing it lacks is a strategy → pillar → epic chain above it.

Nothing that protects correctness is relaxed. What is skipped is a charter the work was never going to earn.

Exit criteria (binary)

This epic has no exit. It has an eviction rule instead:

The measurement that justifies it

Report the share of closed items that came through intake rather than the cascade, at every semi-annual pass. Read a high ratio as evidence against the cascade, not against the people using it: if most real work bypasses four planning layers, those layers are not earning their cost and should be trimmed rather than enforced harder.

This repo currently has no mechanism that can tell it a planning layer is not worth its cost. It took an outside auditor to notice that nine of sixteen conventions had never been used. The intake ratio is the first signal this project has that points the other way — at the corpus rather than at the work.

Out of scope

A second item format, or a lighter Definition of Done. Both would recreate the exact drift the v1.32.0 audit spent eleven findings repairing. If the gates can be skipped for urgent work then they were never gates, and urgency is already carried by Priority: P0.

Publishing focused mode to methodology/. Not until it has been used here. See FUTURE.md for the design and the promotion trigger.

Declared deviation: the WIP cap

E00 is active permanently, so on a literal reading it occupies one of the two WIP slots forever — which would leave a single slot for all real work.

Recorded as a deviation rather than fixed by re-reading the rule: a standing intake epic is exempt from the WIP cap, because the cap exists to limit concurrent chartered effort and intake is not chartered effort. This is a real design finding produced by the first hour of dogfooding, and it is exactly the kind of thing that would have gone unnoticed had the convention been published first and used later.

Linked docs

Risks