test: retry recursive teardown removals — kill a class of phantom failures

[leeovery]

Summary

• A full-gate run went red on ENOTEMPTY inside an afterEach rmSync, in a suite unrelated to the change being made, while four background agents were churning temp directories. The same suite passed 3/3 in isolation.
• Cause: the engine spawns git subprocesses; git drops lock/temp files a moment after the command returns, so a recursive remove that has just emptied a directory then fails to remove the directory itself. Pure timing — it only shows up under parallel load.
• Every recursive teardown removal in tests/scripts now passes maxRetries: 5, retryDelay: 100, which is what the prose-test world builder already used for exactly this reason. 64 call sites across 25 suites, mechanical, no behavioural change.

Test plan

• npm test green: 1699 tests, 0 fail.
• The originally-failing suite re-run 3× in isolation before the change: clean each time, confirming a race rather than a logic fault.

🤖 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 👈 current
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
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