📖About

MEASURED, not suspected. Across the Orion bypass execution book, 1,633 costed runs carry only FOUR distinct values: 0.56 (x1431), 0.10 (x116), 0.03 (x50), 0.20 (x36). Those are exactly the entries of the COST_ESTIMATE table in the bypass module - code 0.56, review 0.10, test 0.10, analysis 0.03, default 0.20. So the execution book's cost_usd column holds ESTIMATES, and the initiatives API sums that column and answers it in a field named actual_cost_usd. An identical figure on five initiatives of wildly different sizes is not a measurement.

Meanwhile the SHARED COST LEDGER, where every real model call appends its cost and tokens against an initiative id, holds distinct and believable figures for those same initiatives: 0.0945, 0.1004, 0.0998, 0.0910, 0.1307, 0.0983, 0.1018.

ONE QUESTION, THREE DERIVATIONS (cookbook 1126): the bypass closer reads the cost ledger; the initiatives API sums droid_runs across the three execution books; and the actuals roll-up sums worker_outputs, which a bypass never produces, so it computes 0.00 for every bypass ever run. Whichever answers first silently wins, and today two initiatives closed hours apart read 0.56 and None while the ledger knew the real number for both.

BUILD: ONE derivation of what an initiative cost, in one module, reading the ledger that actually records spend. Estimates stay estimates - they may be shown AS an estimate, never summed into an actual. Unknown answers None, never zero and never the estimate. Both API endpoints and the actuals roll-up call that one derivation, so the question cannot be derived a second time and drift.

⚖️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)
• [delivery-check] 5 pending must-have criteria stamped met at bypass_end on live re-performance evidence (CAROL-INI-3020): test test_ini3894.py: PASS (10 passed in 5.26s) (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

• One module answers what an initiative cost, reading the shared cost ledger, and returns the tokens with it and the source of the figure. (must_have)
• No surface presents an execution-book estimate as an actual: a regression proves that no reported actual equals an estimate-table constant purely because that constant was stored on a run. (must_have)
• The initiatives API's list and detail endpoints, and the actuals roll-up, all answer from that one derivation and agree with each other for the same initiative. (must_have)
• A bypass initiative reports its true ledger cost and non-zero tokens instead of 0.00 or None - proven on the initiatives closed this session, whose ledger figures are known. (must_have)
• An initiative with no recorded spend reports unknown, never zero and never its budget. (must_have)