📖About

Restore Carolverse innermost self-healing loop (Albus). Root cause: Albus troubleshooter wake is correctly wired into the step reviewer, but the step reviewer only runs after a step tasks report evidence in sized_tasks (0 rows for code steps; Forge records proof in executions). So code steps never get a reviewable result -> reviewer never runs -> Albus never woken -> step strands and blocks unfixed. Fix BOTH: (1) evidence-capture: code/G-type steps produce a reviewable result by reading Forge proof from executions (mirror the skill-step rescue) so the step reviewer runs and wakes Albus normally; (2) strand/fail fallback: when a step strands or fails for ANY reason (incl no-evidence), directly wake Albus troubleshooter BEFORE blocking, so Albus is never starved. Albus must NEVER be invoked on already-blocked initiatives (his realm is active/reviewing only).

⚖️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)
• [status-router] executing -> reviewing | event=bypass_reviewing | bypass transition (or-bx-01)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion assumed the dispatcher was always available, which is false. The initiative spun in 'executing' with no progress for >=10 minutes because there was no health gate on the execution engine itself. Adding this criterion ensures that the strategic plan does not schedule a step that cannot be executed, breaking the deadlock loop. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion assumed the dispatcher was always available, which is false. The initiative spun in 'executing' with no progress for >=10 minutes because there was no health gate on the execution engine itself. Adding this criterion ensures that the strategic plan does not schedule a step that cannot be executed, breaking the deadlock loop. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion required the reviewer gate to fire as a precondition for Albus wake, creating a circular dependency. The fix decouples detection from the evidence pipeline: Albus watches initiative status directly (executing+stale), not through the reviewer's evidence chain. This breaks the deadlock and allows the innermost loop to function. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion required the reviewer gate to fire as a precondition for Albus wake, creating a circular dependency. The fix decouples detection from the evidence pipeline: Albus watches initiative status directly (executing+stale), not through the reviewer's evidence chain. This breaks the deadlock and allows the innermost loop to function. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'Restore Albus innermost loop' is open-ended with no measurable success gate, causing infinite cycling. Replacing it with a bounded criterion (one confirmation then terminate) stops the loop and lets the dispatcher move on. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'Restore Albus innermost loop' is open-ended with no measurable success gate, causing infinite cycling. Replacing it with a bounded criterion (one confirmation then terminate) stops the loop and lets the dispatcher move on. (elrond)
• [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

• Code/G-type steps produce a reviewable result (reviewer reads Forge evidence from executions when sized_tasks is empty) (must_have)
• Elrond step reviewer runs for code steps and wakes Albus troubleshooter on failure (must_have)
• When a step strands or fails for any reason, Albus troubleshooter is woken before the initiative blocks (must_have)
• Albus is never invoked on an already-blocked initiative (active/reviewing realm preserved) (must_have)
• On a real code-step failure, Albus actually fires and produces a verdict/fix (verified) (must_have)