📖About

Ninad's instruction (CLI-171, 2026-07-30). An agent perceives its initiatives, its apps, its services and its own self-model — but never whether the droids it dispatched actually ran. So it reads silence as breakage.

This is the single root cause of seven of the twelve temple wishes. Heimdall asked for his 'quarantined' pipeline to be repaired while his credential rotation was running 14 of 14 clean. Inspector reported his checker had 'gone dark twice in a row' while it ran 4,032 times without a failure. Hermione filed for two dark wake lanes that had each woken 29 times. The runs were all recorded; no agent could see them.

Give every mind a plain account of its own droids on each wake: when each last ran, whether it finished, and — the part that matters — a VISIBLE record when a droid has not run at all, so 'ran and found nothing' can never again be confused with 'never ran'.

⚖️Decisions

• Elrond's bypass methodology checklist (a reminder, not a gate -- you've got this): 0. File it requested_mode='bypass' (planner-vs-bypass is a deliberate choice). bypass_start REFUSES a non-bypass initiative (CAROL-INI-1846), and the dispatcher only skips the bypass lane when the mode says bypass -- a 'planner' mistag lets Merlin's pipeline grab the placeholder step and block your finished work. 1. Filed as planned status -- let the bypass claim/activate it; never file active. 2. Open the bypass (bypass_start) with your droid id + the remediation answer (remediates_initiative_id=NNN, or remediates_nothing=True). 3. Work the blocks for your work-type: template -> design -> code -> test -> review. Do the real work; record decisions on the initiative as you make them. 4. Reality is recorded for you at close -- code (files changed), each decision, and the twin-review verdict become real activities tied to this initiative and show in the Activity Tracker like a planner run (CAROL-INI-1840). No dummy rows. 5. Keep the initiative status moving; it parks in 'reviewing' and is tagged uat-pending for you at close (CAROL-INI-1836), so the stuck-watchdog leaves it alone until UAT. 6. Close runs the gates (design/architecture compliance + caller-audit). If a gate flags something pre-existing or unrelated to your change, waive it with a clear written rationale -- audit, don't skip. 7. Bypass skips the planner's auto-orchestration, NOT the standards. Same template checklist, same review, same observability as a planner run. (elrond)
• [status-router] planned -> executing | event=bypass_executing | bypass transition (or-bx-01)
• An agent now reads its own run record before concluding a lane is broken — New accessor derived entirely from the registry and the run book, injected into what a mind perceives. A newly registered droid appears with nothing edited. Fail-open: an unreadable store yields an empty block and the agent proceeds exactly as before. (orion)
• The account says which it is, in words, for every droid the agent owns. (orion)
• DEFECT IN MY OWN WORK, caught before shipping by measuring — The first cut was unbounded: Elrond owns 46 droids and his block measured 7,018 characters against the under-4,000 compact-context rule — the SAME defect the previous session shipped and had caught for it by the existing check. Measured across all 42 agents before wiring it up. Now bounded to 2,400 with the remainder counted rather than silently dropped. (orion)
• SECOND defect in my own work: the ordering buried the answer — Sorting every never-run droid to the top filled Elrond's list with twelve event-triggered droids resting normally, while the mind and workers that DID run were rolled into a footnote — the opposite of what he needs. A TRIGGERED droid with no runs is its normal state; a SCHEDULED one that never ran is a real alarm. Ordering now distinguishes them, and the summary line stopped calling 30 resting droids 'in need of attention'. (orion)
• It immediately surfaced a real fault nobody had reported — With the account in place Elrond can see his Daily Planner has run 7 times in 7 days and FAILED every one, and that three of his scheduled droids have never fired at all. None of that was in any wish; it was simply invisible. (orion)
• NOT MINE: test_ini2905's 'marker dedupe' failure — The three-layer context check has one failing assertion. It does not import the modules changed here, and the module it exercises was last written at 20:06 — before this session's first edit at 21:06 — by the parallel session still executing the common-prompt work. Reported, not chased. (orion)
• [status-router] executing -> reviewing | event=bypass_reviewing | bypass transition (or-bx-01)
• [status-router] reviewing -> closed | event=operator_signoff | Auto-accepted (CAROL-INI-1859): Orion-initiated, >2 days in reviewing with no objection. (el-srac-01)

✅Success criteria

• An agent waking up can say when each of its own droids last ran and whether it finished. (must_have)
• A droid that has never run, or has stopped running, is stated plainly rather than left silent. (must_have)
• The account is derived from the run records, so a new droid appears without anyone editing anything. (must_have)