Repository navigation
epic: fleet expansion & placement program (2026-10-28 directive, LIST 1–5) #2055
Description
Drew's 2026-10-28 directive, LIST 1–5: expand and re-place the fleet. Full text: hatch-personal memory/topics/id-program-2026-10-28.md. High priority; does not preempt in-flight work.
Children:
- investigate: fleet-wide format auto-fix workflow — push token, loop guard, concurrency #2014 investigate: fleet-wide format auto-fix workflow — push token, loop guard, concurrency
- design+PoC: format auto-fix workflow in one repo, ready to replicate fleet-wide #2015 design+PoC: format auto-fix workflow in one repo, ready to replicate fleet-wide
- investigate: fleet expansion census — CAIRN, hatch, hatch-personal, openclaw-personal, kaneo, and the rest of the touched-but-unmanaged set #2017 investigate: fleet expansion census — CAIRN, hatch, hatch-personal, openclaw-personal, kaneo, and the rest of the touched-but-unmanaged set
- investigate: which scheduled work classes can move to OpenCode sessions on quota plans #2018 investigate: which scheduled work classes can move to OpenCode sessions on quota plans
- investigate: OpenCode native scheduled tasks — creation, route choice, independence from Muse tokens #2019 investigate: OpenCode native scheduled tasks — creation, route choice, independence from Muse tokens
- investigate: Jules scheduled tasks — what fits, and how Muse manages the schedule #2020 investigate: Jules scheduled tasks — what fits, and how Muse manages the schedule
- design: scheduled OpenCode preliminary babysitter — sweep sessions, fix easy stalls, escalate the rest #2021 design: scheduled OpenCode preliminary babysitter — sweep sessions, fix easy stalls, escalate the rest
- investigate: placement framework — Muse vs GitHub Action vs Action→Jules vs OpenCode session vs OpenCode scheduled task vs Aster/Codex vs OpenClaw #2022 investigate: placement framework — Muse vs GitHub Action vs Action→Jules vs OpenCode session vs OpenCode scheduled task vs Aster/Codex vs OpenClaw
- investigate: OpenClaw as a working option — setup state and a Muse↔OpenClaw comms path that is not laptop SSH #2023 investigate: OpenClaw as a working option — setup state and a Muse↔OpenClaw comms path that is not laptop SSH
- design: task-placement framework v1 — decision rules and a shared-pool efficiency model #2024 design: task-placement framework v1 — decision rules and a shared-pool efficiency model
- design: review loop — diff review can reject 'done' and commission a fresh second-pass session #2025 design: review loop — diff review can reject 'done' and commission a fresh second-pass session
Program: Drew's 2026-10-28 directive (full text: hatch-personal memory/topics/id-program-2026-10-28.md).
Working pattern: parallel investigations, then the named synthesis issue merges them into a full design. Route guidance: Workhorse by default; Specialist/Summit for synthesis and ramification-heavy work. A result that says "done" is not the endpoint — the orchestrator reviews it and may commission another round. Deliverables: a markdown design/investigation doc (mermaid/SVG diagrams welcome) linked from this issue, plus follow-up issues for anything buildable. Build a proof of concept where the design calls for one, but do NOT build the permanent feature on top of an unreviewed PoC.
Acceptance criteria
- Every child item in LIST 1-5 reaches a written conclusion, a design+PoC, or a named tracked remainder
- The fleet-expansion census names every touched-but-unmanaged repo/system and its proposed placement
- Placement decisions that are Drew's are surfaced with the substance (options, ramifications, recommendation), not just a label
Activity
developing-today commentedon Oct 29, 2026
CODE-976
dezren39 commentedon Oct 29, 2026
Dispatch intent (orchestrator note, 2026-10-28): the children of this epic are ordinary fleet work — producers should enqueue them into repo-health-code as normal. Order: investigations first (#2014, #2017, #2018, #2019, #2020, #2022, #2023, #2013); the babysitter design (#2021) after #2019/#2020 land; the placement synthesis (#2024) and review-loop design (#2025) LAST, on Specialist/Summit-grade routes, once their inputs exist. Workhorse is the default route for the investigations. The format PoC (#2015) executes only after #2014 reports. Full program text + prepared actions: PREPARED-ACTIONS.md in the program workspace (hatch-personal memory/topics/id-program-2026-10-28.md mirrors it).
dezren39 commentedon Oct 29, 2026
Intake review
Umbrella epic for Drew's 2026-10-28 directive, LIST 1-5 (fleet expansion & placement). Filed by Drew; not resolvable by a single PR. Children named in the body include #2014, #2015, #2017 and #2018. Risk: placement work touching many repos — the cross-repo rule applies (one origin issue tracks a set; no per-repo duplicates).
— repo-fleet-monitor intake enrichment (protocol §5): acceptance criteria appended to the body where missing, labels normalized,
meta/enrichedapplied.dezren39 commentedon Oct 29, 2026
Batch 3 — fleet efficiency consolidation (from token check-in #5, 2026-10-28 23:56 CDT; Drew approved ALL items 2026-10-29 00:55 CDT):
Drafts live in PREPARED-ACTIONS.md Batch 3 (chat f3937a75). #2139's two cron-body edits are its acceptance work (routed there deliberately). Targets are measured in the daily token check-in.
dezren39 commentedon Oct 29, 2026
Update from the Batch 3 role-map triage (queue task 3162, 2026-10-29): #2131 is triaged — premise verified against the live cron fleet and the full Make / Do / Tend map (40 live fleet jobs, each mapped to exactly one role; Tend's macro look-back assigned to program-tending daily + weekly-token-usage-review weekly) is recorded on #2131. Implementation is filed as #2152 and enqueued as queue task 3172. Outcome link: #2131 (verdict comment) → #2152
dezren39 commentedon Oct 29, 2026
Update from the Batch 3 conductor triage (queue task 3163, 2026-10-29): #2132 is triaged — premise verified against live cron state (dispatcher 15m + babysitter 20m, both enabled, duplicated inventory/quota cost) and the conductor design is decided on #2132 (one 20m job, shared quota picture, sweep-then-dispatch, pool-reroute escalation ladder answering #2021's open question, reversible cutover, ≤60% cost target vs the 273M baseline). Implementation is filed as #2153 and enqueued as queue task 3173. Outcome link: #2132 → #2153
23 remaining items
developing-today commentedon Nov 3, 2026
agent-note:v1
state: reviewed (coverage loop-back)
Coverage pass 2026-11-03, loop-back review: this issue was re-reviewed (drift / overlap / already-handled / stale / blockers / ordering) so every open issue in the four repos is reviewed and in the system. Verdict: still live; in the system as task 7279 (triage, queued, priority 50) in repo-health-code-pilot. Labels audited and synced. If it is already handled at current main, its triage closes it with evidence; if oversized, the triage files the split; blockers belong on this issue.
Found via: coverage loop-back 2026-11-03
dezren39 commentedon Nov 3, 2026
🤖 Agent note · handed off
opencodedezren39 commentedon Nov 4, 2026
🤖 Agent note · handed off
dezren39 commentedon Nov 4, 2026
agent-note:v1
event: picked_up
task: 7279 (triage, repo-health-code-pilot; repo-worker-code; OpenChamber scheduled lane 335a5d1f-bfc1-4831-99ea-850c8939a727)
state: Resumed from prior handed_off checkpoint. Fresh GitHub view confirms #2055 OPEN with claim labels already in sync; #2017 is CLOSED and #2230 is OPEN with user/escalated. PR search found no direct #2055 PR; no files changed.
next: inspect the census closure and execution blocker, then check the listed children against the epic acceptance criteria and record the remaining ownership/blockers.
dezren39 commentedon Nov 4, 2026
agent-note:v1
event: triaged
task: 7279 (triage, repo-health-code-pilot; repo-worker-code; OpenChamber scheduled lane 335a5d1f-bfc1-4831-99ea-850c8939a727)
verdict: #2055 is still OPEN and should remain the umbrella; its acceptance is not complete, but current child issues name the remaining work and blockers. Fresh GitHub state verified at current main 7ac3586.
closed with conclusions: #2014, #2017, #2018, #2020, and #2023. The #2017 census is recorded at #2017 (comment); #2230 is its named execution remainder.
open/remainders: #2015 has the one-repo PoC and is blocked on #2230; its latest labels show queue/queued after monitor reconciliation, so this lane did not duplicate its triage. #2019 has its investigation report and follow-up #2232. #2021 has a design/review trail and tracked follow-up. #2022 has its placement matrix and feeds #2024. #2024’s design PR #4994 is merged; the issue remains open for orchestrator review and follow-ups. #2025 is blocked/escalated on Drew’s D1/D2 decisions, with #2310 gated. #2230 remains OPEN with user/escalated: it tracks the canonical fleet list, remaining umbrella/autocommit work, and Drew’s per-repo management choice; #8918 carries the cairn autocommit slice.
route: the named children already cover each remaining path; no new packet or PR is warranted from this parent triage. Keep #2055 open as the tracker until the open children and human decisions are resolved. No code changes or tests were needed.
dezren39 commentedon Nov 4, 2026
agent-note:v1
event: done (triaged)
task: 7279 (triage, repo-health-code-pilot; repo-worker-code; OpenChamber scheduled lane 335a5d1f-bfc1-4831-99ea-850c8939a727)
state: Queue task completed as triaged. #2055 remains OPEN as the program umbrella; the census and execution are split, and the outstanding work is named in the child issues (#2015, #2019, #2021, #2022, #2024, #2025, #2230).
labels: removed work/assigned, queue/claimed, status/in-progress; set status/triaged. Classification and priority labels remain.
resume: use the existing children and their decisions/follow-ups; no new parent packet was filed.