Summary

• Revives a hook that has never fired. prose-walker declares a stop hook, but record-action.cjs scopes every event to a world by finding the world path in the payload — and a stop event carries no tool input, so there was nothing to match. The record was silently short. Its world now resolves from the agent transcript the payload points at.
• Names the model each walk ran on. The same transcript states the model per message. An edited agent definition doesn't reach a running session until plugins are reloaded, so a walk can run on a model nobody intended and every verdict is then trusted at the wrong one. The model travels from the harness record → asserter → orchestrator verdict block, so a stale definition shows up in the output instead of being something to remember.
• Hardens the asserter-violation log — it appended to a path whose parent it never created, so a contract breach went unrecorded wherever that directory was absent.
• Drops a stale claim in the orchestrator description that it escalates failures to Opus; it hasn't since both agents moved to the same model.
• Adds tests/scripts/test-prose-record-action.cjs (14 tests, registered in npm test) — the recorder had no suite, which is how the dead hook stayed invisible. Covers both world-resolution paths, model extraction (single, multiple, unreadable, partial trailing line), the asserter contract, and the never-break-what-it-observes guarantees.

Test plan

• node --test tests/scripts/test-prose-record-action.cjs — 14/14
• node --test tests/scripts/test-prose-cases.cjs tests/scripts/test-prose-snapshots.cjs — 35/35 with the new suite
• Hook driven end to end with a synthetic SubagentStop: resolves the world from the transcript, extracts the model, tolerates a truncated line, writes one uniform row
• Verified live against a real walk — the payload's agent_transcript_path proved a walk ran on claude-opus-5 while the definition on disk said sonnet, which is the case this change makes visible

🤖 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 👈 current
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
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