T03: correct stale targets in the workplan ($92.87, AC-1)

T05 still named $248.46 as its acceptance test and the Purpose total
still omitted the subagent tree. Both corrected in place, with the
error sequence recorded rather than overwritten.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
tegwick 2026-07-31 08:46:36 +02:00
parent b96cd94a64
commit 900467017f

View file

@ -39,7 +39,9 @@ repeatable acceptance target:
| Cache read | 80,453,702 | **$59.59** | | Cache read | 80,453,702 | **$59.59** |
| Cache write (1h) | 1,672,854 | $21.95 | | Cache write (1h) | 1,672,854 | $21.95 |
| Input | 676 | $0.00 | | Input | 676 | $0.00 |
| **Total** (339 responses: 206 opus-5, 118 fable-5, 14 sonnet-5) | | **$92.21** | | Main transcript (339 responses: 213 opus-5, 118 fable-5, 14 sonnet-5) | | $92.21 |
| Subagent tree (adversarial review, ran inside the pin) | | $0.66 |
| **TRUE TOTAL** | | **$92.87** |
The headline finding survives the correction and gets sharper: The headline finding survives the correction and gets sharper:
**88.4% of the cost is cache, against 314k tokens of output — a 256:1 **88.4% of the cost is cache, against 314k tokens of output — a 256:1
@ -166,10 +168,16 @@ state_hub_task_id: "bea4cc0a-e4d0-4077-9dc7-df7726a48f86"
``` ```
Run the collector over the CB-WP-0001 session and commit Run the collector over the CB-WP-0001 session and commit
`evidence/CB-EV-0002-cost-accounting.md`. Reconciliation against the `evidence/CB-EV-0002-cost-accounting.md`. The acceptance test is
totals in this workplan's Purpose section is the acceptance test: the `specs/CostAccounting.md` AC-1: reproduce **$92.87 pinned at `fc76445`,
tool must reproduce $248.46 (Fable 5) from the same transcript, or as its two components** ($92.21 main transcript + $0.66 subagent tree),
explain the difference. plus AC-2's zero reconciliation residual.
*(Originally written as "must reproduce $248.46". That figure was wrong
by 2.7×, and the first corrected target — $92.21 — was itself reachable
only by a collector carrying the subagent blind spot the survey had just
documented. Both errors are recorded rather than quietly overwritten;
the sequence is the point.)*
Then use it to answer at least one question that could not be answered Then use it to answer at least one question that could not be answered
before, and record the answer. Candidates: which of T01T09 cost most before, and record the answer. Candidates: which of T01T09 cost most