fix(detect): respect cache root for converted sidecars

[hopstreax]

Summary

Fixes #2787.

detect() previously wrote converted Office document sidecars to the scanned source tree even when cache_root was provided. This could mutate read-only or pinned source checkouts during detection.

This change makes converted sidecars respect cache_root while preserving the source root as the hashing/identity anchor.

Changes

• Redirect converted sidecars to <cache_root>/graphify-out/converted/ when cache_root is provided.
• Preserve the existing <root>/graphify-out/converted/ behavior when cache_root is not provided.
• Continue passing the original source root to conversion helpers so sidecar filenames remain stable.
• Added a regression test verifying that:
  • the scanned source tree is not polluted with graphify-out/
  • the converted sidecar is written under cache_root
  • detect() returns the redirected sidecar path
  • the sidecar content is generated correctly
  • the sidecar hash remains anchored to the source root

Validation

• uv run pytest tests/test_detect.py -k "detect_office_conversion_respects_cache_root or convert_office_file_sidecar_name_stable_across_checkouts" — 2 passed
• uv run pytest tests/test_office_incremental.py tests/test_office_limits.py tests/test_google_workspace.py tests/test_extract_cache_location.py — 22 passed
• uv run pytest tests/test_detect.py — 235 passed, 11 skipped

[graphify-labs]

Graphify reviewed this change.

Worth a look — the grounded gate found no coupling regressions or blocking issues, but 1 advisory finding(s) below merit a look before merge.

Formal verification. No changes could be formally verified in this run.

---

Graphify review — findings

This PR modifies detect() in graphify/detect.py so that when a cache_root argument is provided, converted document sidecars are written under cache_root/GRAPHIFY_OUT/converted instead of under the scanned root directory. It adds a new test verifying that the scanned corpus tree is left untouched, the sidecar is placed under cache_root, and the sidecar filename hash remains anchored to the scan root. The large list of changed test/rationale symbols appears to reflect line-number shifts from inserting the new test rather than substantive changes to those cases.

Worth a look

• cache_root parameter may not exist in detect() signature — graphify/detect.py:1638 · Escalate · high
  • agreed by 2 of 2 members but NOT verified (no proof, no reproducing execution) — consensus is not a verdict; needs human review Execution auto-disposal is off for this run; enable it (with sandbox isolation) to have Graphify try to confirm or refute this automatically.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 1527 functions depend on the 482 functions this change touches.

Health — this change adds coupling hotspots:

• new: extract() — 463 callers, 41 callees
• new: _rebuild_code() — 95 callers, 51 callees
• new: detect() — 100 callers, 15 callees
• new: save_manifest() — 34 callers, 11 callees
• new: extract_files_direct() — 17 callers, 20 callees
• new: extract_xaml() — 19 callers, 17 callees
• new: extract_corpus_parallel() — 26 callers, 10 callees
• new: dispatch_command() — 2 callers, 117 callees
• …and 25 more

Verification — 1527 functions in the blast radius were not formally verified this run (proofs are advisory here).

Gate & verification

graphify gate

PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.

Advisory (not blocking):

• verification_scope: 785 function(s) in the blast radius were not formally verified this run

Formal verification

Could not verify: Could not verify detect.

The verifier did not have enough to check detect, so it is saying so rather than guessing. No false assurance is the whole point.

Guarantee: No guarantee either way, this is an honest abstention, not a pass.

Note: Reason: parameter `root` is annotated `Path` — outside the synthesizable primitive/collection set

· 33 more finding(s) on lines outside this diff (see the check run).