Summary

Found by deliberately breaking a skill to prove the harness catches it. It did — and turned up something else on the way.

The orchestrator returned:

MODEL: first walk claude-sonnet-5, escalated rerun claude-opus-5
       (asserter: claude-sonnet-5 both times)

The asserter was not on Sonnet. Every asserter transcript from that run is claude-opus-5 with zero tool calls, exactly as contracted; the Sonnet agents were the orchestrators, which are Sonnet by design.

And it had no way to know either way. The only model anywhere in the record comes off the walker's SubagentStop line. No agent's own model appears anywhere it can read.

So a slot that exists to carry a fact carried a guess — in the one field whose whole purpose is catching a run that silently used the wrong model. MODEL: now states the walking model and nothing else.

The break test itself (reverted, never committed)

Moved render phase-note above the status read in workflow-investigation-entry — a plausible "announce the phase first" reordering. Result:

VERDICT: FAIL
FAIL  calls_in_order — "render phase-note" never ran after
                       "manifest get crash-fix.investigation.crash-fix status"
PASS  calls_include  — ran all of: … status, render phase-note
N/A   engine_before_write — no workflow state was written
PATH: 4/5 — step 4 fails on both runs

Every layer behaved:

• calls_in_order caught it while calls_include passed — both calls ran, just out of order. Precisely the gap ordering was added for, now proven on real prose rather than a unit test.
• engine_before_write reported N/A, not a misleading green tick.
• Escalation fired for the first time: Sonnet walked and failed, the orchestrator rebuilt the world and reran on Opus, Opus failed identically → confirmed defect rather than flake.
• The CHECKS line surfaced the computed facts instead of leaving them to be inferred from a green overall verdict.

Test plan

• Prose gate unchanged and green
• Working tree verified clean after revert; the broken prose was never committed

🤖 Generated with Claude Code

Stack

1. docs(design): prose-tests programme design log #544
2. feat(prose-tests): the framework — cases, worlds, runner, skill #545
3. test(prose): feature happy-path corpus — five worlds, seven cases #546
4. test(prose): bugfix corpus — the investigation-centric surfaces #548
5. test: retry recursive teardown removals — kill a class of phantom failures #549
6. fix(entry-skills): close the handoff fences — six files render their arms wrong #550
7. docs: a contributing page for working on the system #551
8. fix(entry-skills): every handoff arm says to invoke the skill #552
9. fix(implementation): environment setup belongs to the setup reference alone #553
10. fix(prose-tests): the asserter is told which substitutions were armed #554
11. feat(prose-tests): the mid-flow substitution, and a world only prose can describe #555
12. test(prose): claims assert consequences, not what was displayed #556
13. feat(prose-tests): record everything the agents do, results included #557
14. fix(discussion-entry): the handoff reports the source it actually had #558
15. fix(prose-tests): the stop hook records, and names the model that walked #559
16. fix(prose-tests): command output was never actually recorded #560
17. feat(prose-tests): judge the walk as told, not the summary returned #561
18. feat(prose-tests): decide in code what an agent should not be deciding #562
19. test(prose): a case starts where a session starts #563
20. feat(prose-tests): walk on Sonnet, judge on Opus, escalate a failure #564
21. test(prose): give the eight read-only cases something that can fail #565
22. test(prose): only walks that can be observed, and checks that survive the trip #566
23. fix(prose-tests): the verdict names only the model the record names #567 👈 current
24. test(prose): discovery, walked to the point where work first exists #568
25. fix(prose-tests): the asserter judges which of prose or walker was at fault #569
26. docs(conventions): a step whose reference routes every exit still signposts #570
27. test(prose): discovery's epic arm, to the same durability boundary #571
28. fix(prose-tests): keep the walk's closing turn, and name the one live difference #572
29. fix(prose-tests): prescribe the cross-check, drop a check that cannot fire, lint the scope #573
30. test(prose): declare the prose each case actually walks #574
31. fix(investigation): the symptom interview never ran for work shaped in discovery #575
32. feat(prose-tests): conduct — playing a person where a script cannot reach #576
33. fix(prose-tests): what a write put in a file is evidence, not incident #577
34. fix(investigation): the interview starts from the carrier, not over it #578
35. test(prose): quick-fix scoping, the whole definition stage in one walk #579
36. fix(prose-tests): the world log is the walker's record #580
37. fix(prose-tests): a confirmed failure reports both runs' checks #581
38. fix(prose-tests): each prescribed command runs as written #582
39. fix(scoping): context gathering reads the whole carrier, asks only the gaps #583