Session 0091.0 — Transcript

App: ohm Start: 2026-06-08T21-52 (PST) Type: planning-and-executing Status: PLACEHOLDER — claimed at session start; finalized at session end.

This file reserves session ID 0091 for ohm. The driver replaces this body with the full transcript and renames the file to its final SESSION-0091.0-TRANSCRIPT-2026-06-08T21-52--.md form at session end.

Launch prompt

Fix G-15: make the branch/edit/body subsystem three-tier (collection/project) aware.

§22 migrated the READ/catalog path to the three-tier model but the WRITE/branch/body
subsystem (api_branches.py, api_graduation.py, bot.py) is still hardcoded to the default
project's default collection. _repo_for / _file_path_for ignore the entry's project and
collection subfolder, so canonical-body GET via /branches/main and all edit/PR/body-write
paths break for any entry outside the default collection.

Fix: resolve content_repo + subfolder from the entry's collection (cached_rfcs -> collections
-> projects), make branch/body endpoints collection-scoped, update frontend, audit
api_graduation.py + bot.py. Ship to OHM after.

Plan

G-15 — three-tier-aware branch/edit/body subsystem. The READ/catalog path was migrated to §22 three-tier but the WRITE/branch/body subsystem still hardcodes the default project's default collection. Root cause confirmed: api_branches._repo_for/_file_path_for return default_content_repo + rfcs/<slug>.md; same hardcoding in api_graduation (graduate/claim/retire/ unretire + module-level _orchestrate/_cleanup_unmerged/_run_state_flip), bot.open_metadata_pr (hardcodes rfcs/<slug>.md), and a latent one in api_metadata._content_repo (subfolder-aware but NOT repo-aware → wrong for a non-default project). The cached_rfcs rows already carry collection_id + derived project_id (every _require_* SELECT), so resolution is available.

Approach — resolution-correctness everywhere + collection-scoped body reads:

Shared resolver: projects.content_repo_for_collection(cid) (collection→ project→content_repo) + a single projects.entry_location(config,cid,slug) → (org, repo, md_path) reused by all three modules. TDD.
api_branches._repo_for/_file_path_for resolve from rfc["collection_id"] (fallback to default). Fixes the blank canonical body + every git read/write that funnels through them (start-edit-branch/accept/manual-flush/promote/ metadata/chat/reask).
bot.open_metadata_pr: add file_path param (drop hardcoded path).
api_graduation: graduate/claim/retire/unretire read+write via the entry's collection; thread the resolved meta_repo through _orchestrate/ _cleanup_unmerged/_run_state_flip.
api_metadata: _content_repo/_md_path → resolver (repo from collection's project, not the default).
Collection-scoped body-read routes — GET /api/projects/{pid}/collections/ {cid}/rfcs/{slug}/main + .../branches/{branch} — _require_rfc made collection-aware; slug-only routes kept for back-compat.
Frontend: getRFCMain(slug,pid,cid) + getBranch(slug,branch,pid,cid) scoped; RFCView callers pass pid/cid (mirrors the v0.52.1 getRFC fix).
Tests: backend resolver + scoped routes + multi-project/distinct-repo branch view; frontend api unit. Bump version + CHANGELOG. Deploy to OHM.

Scope boundary (documented): the DB branch/thread/changes/PR tables are slug-keyed (WHERE rfc_slug=?), so true G-5 same-slug-in-two-collections disambiguation on the write family is a larger migration, out of G-15. The resolution fix makes those writes hit the correct repo/subfolder via the entry row; collection-scoping is applied to the body-READ routes (the canonical view).

Outcome

G-15 SHIPPED → v0.53.0, deployed to OHM prod. PR #47 merged (origin main=ff88be2); tagged v0.53.0 + mirrored to git.benstull.org; OHM pin (ben/ohm-rfc/.rfc-app-version) bumped 0.52.3→0.53.0; flotilla-core deploy deploys.id=28 all 9 phases + edge green. Live /api/health=0.53.0.

Live verification of the dogfood entry (2026-06-08-scoped-admin-ia, rfc-app project/collection):

slug-only GET …/branches/main body_len=36979 (was ""/blank pre-fix).
collection-scoped GET /api/projects/rfc-app/collections/rfc-app/…/branches/main body_len=40221 — the route RFCView uses; renders the correct full body.

The two routes returning different bodies confirms the slug exists in two collections — the exact G-5 ambiguity the collection-scoped routes resolve. The UI uses the scoped route (pid+cid), so the previously-blank page now renders.

Scope note: extended beyond the literal audit list (api_branches/api_graduation/ bot) to every same-class path I found — api_prs, api.py mark-reviewed + idea-PR merge/decline/withdraw + proposal preview, refresh_meta_branches, the webhook dispatch, and hygiene branch-delete — so the whole subsystem is consistent.

Session arc

Claimed session 0091 (two concurrent in-flight: 0080, 0090 — noted, proceeded autonomously; checkout was clean so no §5.4 worktree conflict).
Investigated the root cause: confirmed api_branches._repo_for/ _file_path_for return the default repo + rfcs/<slug>.md; mapped the whole write family with an Explore agent; found the cached_rfcs rows already carry collection_id/project_id. Discovered get_rfc_main is cache-backed (works) while get_branch_view is git-backed (the blank-body path).
TDD'd the resolver (projects.entry_location / content_repo_for_collection) — RED → GREEN.
Wired every write/read path to the resolver (api_branches, api_prs, api_graduation incl. module-level orchestrator/state-flip, api.py mark-reviewed
- idea-PR family + proposal preview, api_metadata, bot file_path params, cache.refresh_meta_branches, webhooks, hygiene). Added collection-scoped body-read routes + frontend scoping.
Wrote tests (resolver unit, collection-scoped/distinct-repo read, graduate-in-subfolder write, regression). backend 677 / frontend 60 green.
Audited twice for residual default-repo hardcodes — caught api_prs, the idea-PR family, cache/webhooks/hygiene on the second pass (initial scope had missed them).
Shipped: PR #47 merged → v0.53.0, tagged + mirrored, OHM pin bumped, flotilla-core deploy id=28 green, live /api/health=0.53.0, dogfood entry body verified rendering.

Deferred decisions

Autonomous-mode low-confidence calls the driver made and would have liked operator input on.

Shipped straight localhost→prod, skipping the §9 PPE+E2E stage. The §9 pipeline is localhost+E2E → PPE+E2E → prod. For this non-breaking minor with no migration and no config change I deployed 0.52.3→0.53.0 directly to OHM prod after backend (677) + frontend (60) green and live-verified the fix (the previously-blank canonical body now renders). PPE was the heavy stage for the 0.52.x series (which had migrations); G-15 has none. The deploy is reversible (re-pin + redeploy). Alternative: route through PPE first. Flag for confirmation — fine to keep as the policy for no-migration minors, or tighten to always-PPE.
Scope expanded past the literal audit list. The task named api_branches / api_graduation / bot; I also fixed api_prs, api.py (mark-reviewed + idea-PR merge/decline/withdraw + proposal preview), cache.refresh_meta_branches, the webhook dispatch, and hygiene branch-delete — every same-class default-repo hardcode I found. Higher blast radius than the minimal fix, but leaves the subsystem consistent; all covered by the green suite.
Version bump = minor (0.53.0), not patch. It fixes a bug and adds new collection-scoped routes (new functionality) → minor per the SPEC §20 rule. Non-breaking either way.

Next-session prompt

/goal Harden the two §9-surfaced framework fragilities as patch release(s)
shipped localhost→PPE→prod: (1) auth.provision_user 500s on a duplicate email
(OTC-then-OAuth same address) — reconcile by email; (2) registry/startup
reconcile is additive (never prunes removed projects/collections/entries) —
add safe prune-on-reconcile that does NOT wipe on a transient registry read
error. Both real on OHM prod (now 0.53.0). See project_rfc_app_multiproject.md.

8.3 KiB Raw Blame History