fix: skip merge attribute for ignored graph output

[SomSamantray]

Summary

Fixes #2595.

• Keep registering the local merge.graphify driver when graph output is ignored.
• Skip the redundant .gitattributes mutation when Git reports the graph path is ignored.
• Preserve existing behavior for tracked graph output.

Validation

• uv run pytest tests/test_hooks.py -q — 79 passed
• uv run ruff check graphify/hooks.py tests/test_hooks.py — passed
• uv run pyright graphify/hooks.py — 0 errors
• Full suite: 4243 passed, 42 skipped, 15 unrelated baseline failures (missing optional openai dependency and unavailable shallow-clone skillgen baseline refs)

Contributor

Som Samantray

[graphify-labs]

Graphify reviewed this change.

Looks safe to merge — no coupling regressions and no blocking issues, checked against the code graph (not a self-assessment).

Formal verification. 1 change(s) tested, no difference found (not proven).

---

Graphify review — findings

This PR adds a plan document and implements a guard in graphify/hooks.py so that hook installation skips adding the .gitattributes merge rule when the graph output path is ignored by Git, while still registering the local merge driver. It refactors the attribute-line helper by extracting a _merge_attr_path() function and introduces a _graph_path_is_ignored() helper that shells out to git check-ignore. A new test verifies that installation on a repo ignoring the graph output registers the driver but does not create .gitattributes.

No blocking issues surfaced. 2 lower-confidence candidates did not survive cross-model review.

Analysis details — impact, health, verification

Impact & health

Graphify review

Impact — 304 functions depend on the 158 functions this change touches.

Health — this change adds coupling hotspots:

• worse: install() — 26 callers, 6 callees

Verification — 304 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: 165 function(s) in the blast radius were not formally verified this run

Formal verification

No difference found (not proven): No behavior difference found in \_merge\_attr\_line (not a proof).

The verifier ran both versions of \_merge\_attr\_line on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.

Guarantee: Empirical: differential testing (both versions run on many generated inputs). A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.

Note: An input the sampler did not try could still differ.

Could not verify: Could not verify \_register\_merge\_driver.

The verifier did not have enough to check \_register\_merge\_driver, 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

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