Emerging. Elevated 2026-10-13 by CIO under Emerging self-approval (methodology-audit-policy-updates-2026-03-16.md) from PP-002, “Critical vs. Commodity Work in a Role”, tracked in PROTO-PATTERNS.md since 2026-04-27. PM keeps upgrade, revision and removal authority.
Every role has a small core of load-bearing work, where its distinctive judgement lives (a methodology audit, a voice-quality read, an ADR audit, a product-shadow observation). Around that core is commodity work the role also owns: mail, CI checks, logs, deploy steps, trackers. Commodity work gets mechanized (a script, a skill step, a hook, a duty-cycle step) and then runs on schedule. Judgement work can’t become a hook. Unless it has a trigger, it silently stops, and the role still looks busy and healthy from outside.
April (PP-002, by reflection): seven retiring leadership roles independently drew the same load-bearing versus commodity split in their Agent 360 v0.2 section 6 (HOST, CIO, Comms, CXO, PPM, Arch, Exec). The full per-role table is kept in PP-002’s history in PROTO-PATTERNS.md (git log).
October (by measurement, the elevation evidence): docs/internal/research/role-duties-operational-vs-official-2026-10-13.md. That is 11 roles, 164 written duties and 30 days of logs, measured six months after migration.
PP-002’s three elevation criteria, each checked:
<!-- ROLE-CARD --> blocks in each briefing, printed by the SessionStart hook since 2026-10-13) list priority 1 and name the commodity work to keep small.Drift: line in duty-cycle-tick v1.51.A role’s log for a week is all mail, fixes and fire bookkeeping, every entry is correct, and nothing in it is the work the role card names first.
This needs one independent instance outside the 10-13 measurement: the 10-27 re-run, or another project’s team (Klatch was named in PP-002). It also needs evidence that a trigger moved the specialist share, not only that roles acknowledged the gap.