This is the backlog for the methodology project’s self-development (the methodology applied to its own development). Follow methodology/04_backlog_items.md for the canonical BL-#### format, ROI heuristic, lifecycle, and item rules — everything in this backlog uses that format unchanged.
self-development/backlog/
├── README.md # this file
├── EPICS.md # rollup of all epics with status counts + pillar coverage
├── ACTIVE_CONTEXT.md # volatile working-state baton
├── FEEDBACK.md # inbound feedback inbox (untrusted by default)
├── HUMAN_NEEDED.md # items blocked on human-only action
└── epics/
├── E00-intake/ # standing; never closes
├── E01-examples-folder/
│ ├── README.md # charter
│ ├── BACKLOG.md # active items
│ ├── ARCHIVE.md # done items
│ ├── FUTURE.md # deferred items
│ └── TEST.md # epic acceptance + regression scenarios
├── E02-first-semiannual-self-evaluation/ # same shape as above
├── E03-git-workflow-trim/
├── E04-native-tool-templates/
├── E05-cheatsheet/
├── E06-bmad-v6-landscape-pass/
├── E07-agentic-workflow-pass/
├── E08-role-briefs/
├── E09-external-landscape-pass/
├── E10-external-audit-remediation/
└── E11-assurance-and-squads/
The five-file shape is the methodology’s, per “Standard epic-folder structure”.
This diagram documented four for a long time and the filesystem matched the diagram rather than the
methodology — an external audit caught it (F-09). All twelve epics carry all five files. The nine
back-filled TEST.md files are empty-but-present and point at the real verification record in each
ARCHIVE.md; no acceptance scenarios were reconstructed, because back-filling rows that never ran would
be fabricated evidence. The rule was not weakened to fit the practice.
E<NN> (zero-padded). Currently E00 through E11; E00 is the standing intake epic and does not close. Folder names carry the E prefix (E01-examples-folder), per “Standard epic-folder structure” in 03_epics.md.BL-<####> monotonic across all epics in this self-development backlog, and shared with FUTURE.md items so promotion needs no renumbering. Highest assigned: BL-0069 (E00, 2026-10-18).Per the methodology’s project structure convention: item IDs are repo-wide-monotonic within this backlog (so items can move between epics without renumbering and grep BL-0042 is unambiguous within the self-development backlog).
Standard methodology workflow per methodology/04_backlog_items.md. Three project-specific overrides:
methodology/ autonomously. Cycle outputs that suggest methodology changes are surfaced to the maintainer; methodology updates ship as a normal release cycle. The full constraint set is in ../AUTONOMOUS_LOOP.md “Hard constraints.”self-development/ ship as minor or patch versions of the public repo — the work is in the public repo, but it’s project-meta work, not methodology work. The abstract methodology in methodology/ versions independently.EPICS.md “WIP cap note” for reasoning. The cap has still never actually been contended, so there is no evidence for raising it further. E00-intake is exempt — it is standing rather than chartered, and a permanently-active epic would otherwise eat a slot forever. Declared in its charter as a deviation rather than read into the rule.E00 runs it; E00/FUTURE.md holds the design and the condition that would promote it into methodology/. Nothing about it is a published convention yet, deliberately.../strategy/00_master_plan.md — phases, pillar roadmap, doc index.../pillars/ — P1 through P9 capability layers; each epic charter names its primary pillar.../ADOPTION_PROFILE.md — what this repo adopted, adapted and omitted, with reasons. Read it before treating this directory as a worked reference.../RELEASE_EVIDENCE.md — the commands behind every published count, and the line budgets that bind this repo.../../methodology/. They are read-only from this backlog’s perspective unless the maintainer explicitly authorizes a methodology change in a separate release cycle.Bootstrapped on 2026-05-25; the bootstrap completed long ago. As of 2026-09-05: eleven epics, eight done, one parked (E04, will not resume), one active (E10), one standing (E00 intake). Item IDs run to BL-0062. The live rollup is EPICS.md — this section records origin, not current state.
This paragraph was stale by two epics, twenty-two item IDs and a WIP cap when an external audit read it. It is the second-order form of the finding E10 exists to fix: a hand-maintained record drifts, and nothing notices until someone outside reads it. No check will catch the next one — a committed checker was declined — so the only defences are that this section says what it is (origin, not current state), that it points at the file that is, and that ../RELEASE_EVIDENCE.md makes the counts reproducible by anyone who cares to look.