rdmsm4x-changelog-20260825-1317-todo-intake-folder-and-scanner
Summary: Stood up the canonical fleet to-do/intake
folder at rdmsm4x:~/dev/todo/ with a periodic scanner
(scan_todo.zsh + LaunchAgent
com.eastcoastscience.todoscan) that classifies intake
items, relays unambiguous ones append-only into owning projects'
ISSUES.md, and logs every action in a generated table. Requested by Rich
via claude@rdmbair15m5 (fleet bus
20260825-131046-EC786538). Work window: 13:11–13:17 EDT,
2026-08-25.
Scope
rdmsm4x only. Additive — nothing moved, renamed, or deleted; the standing FLEET HALT (2026-08-24 00:46) was respected: this was Rich's direct instruction and touched no existing structure except two append operations described below.
Files created
~/dev/todo/README.md— contract: item format, scanner behavior, reconciliation table (ISSUES.md / DECISIONS-PENDING-RICH.md / inbound/BACKLOG.md all KEPT canonical; todo/ is the intake layer above them, not a replacement)~/dev/todo/scan_todo.zsh(v1.0.0, mode 755) — the scanner~/dev/todo/items/— 4 initial items:20260825-claude-remote-control-boot-service.md(blocked-on-rich, 4 decisions; report at https://dev.dataroo.net/claude-remote-control-service-research.html)20260825-domain-portfolio-decision.md(blocked-on-rich; 15 domains bill 2026-08-28)20260825-replicantdb-plan-disposition.md(relayed → apps/replicantDB)20260823-fleet-agent-registry-integration.md(parked; placement waits on halt release)
~/dev/todo/RELAY-LOG.md— generated routing table (item → project → action → timestamp)~/dev/todo/.state/{relay-ledger.tsv,scan.log}— ledger + log~/Library/LaunchAgents/com.eastcoastscience.todoscan.plist— hourly (StartInterval 3600)- on-change (WatchPaths items/), Background/LowPriorityIO/Nice 10 (box hit load 70 yesterday)
Files appended to (shared-write surfaces, append-only)
~/dev/apps/replicantDB/ISSUES.md— one relayed line under new trailing section## Relayed from ~/dev/todo, tagged[todo:20260825-replicantdb-plan-disposition]~/.agent-coordination/checkins/todo-scanner-rdmsm4x-20260825-131537.md— relay check-in
Commands run / verification evidence
zsh scan_todo.zsh -n -vdry-run: 3 park + 1 relay, as designed- Live run relayed 1 item; second live run relayed 0
(idempotency verified: item status + ledger key
<id>::<project>+ on-disk[todo:<id>]tag in target) plutil -lintOK;launchctl bootstrap gui/501loaded;launchctl kickstartproduced a launchd-driven run at 13:16:28 EDT (verified in scan.log)- Ledger verified real TSV via awk field count (5 fields/row)
- Bug found+fixed during validation: zsh
print -rwrote literal\n/\t; switched toprintf, repaired the one affected ISSUES.md line and the ledger before anything consumed them
Undo
launchctl bootout gui/501/com.eastcoastscience.todoscan,
delete the plist, rm -rf ~/dev/todo, remove the tagged
section from apps/replicantDB/ISSUES.md. No other
state.
Outstanding owner actions (Rich)
- The 4 Remote Control decisions; the 2026-08-28 domain deadline (DECISIONS-PENDING-RICH A1)
- Apple Notes changelog entry PENDING —
notes_changelog.zshis failing fleet-wide ("Notes never became responsive within 600s", see ~/dev/ISSUES.md). File copy (this file) written; Notes entry to be added when Notes automation works again.