test(code-index): assert exact scheduler and snapshot outcomes

[ScriptedAlchemy]

Summary

• Replace weak scheduler, snapshot, import, scope-root, and temporal assertions with the literal outcomes those fixtures produce.
• The every-weight profile digest test now pins both checked-in weights (lexical_weight_ppm 1000000 and graph_weight_ppm 250000) and shows each one moves the digest.
• Annotation edge-count pins in scala, vbnet, and php stay as they are: a name check already sits beside each count.

Mutation proof

Each rewrite failed while the old weak predicate stayed true, then passed after git checkout -- restored production:

• Missing branch: NotFound { branch: "missing" } failed when the error became ReferenceUnavailable (is_err still held).
• Profile digest: zeroing lexical_weight_ppm before the hash failed the lexical inequality.
• Scheduler seat: a second record_injected_activation_attempt made the attempt count 2 (> 0 and is_some still held) and failed the literal count and seated src/main.rs / main function asserts.
• Non-repo roots: renaming the error to git_repository_missing failed the literal git_repository_unavailable (is_err still held).
• Import tamper: ImportEmptyBindingName failed the ImportAuthorityMismatch match.
• Temporal rank: ascending raw_score failed the left-then-right anchor order.

Suites

Non-zero results on the touched crates: code-index lib 260, code_index_suite 182, resident_accounting 3, retention lib 111, runtime lib 545 (1 ignored), query lib 272 (1 ignored), canonical_execution_equivalence 5, index_bench_workload 1, retrieval_contract_spine 2, search_quality_suite 75, temporal-query lib 179, temporal hotpath_coverage 1, search-eval bins 3 and 3. Focused rewrites: 1+1+2+1+1+1 passed. Clippy -D warnings on the touched crates passed. cargo fmt --all -- --check passed.

tracedecay-search-eval lib is 35 passed and 1 failed: baseline_report_retains_raw_fallback_current_and_exact_ten_x_samples compares query-fallback train sha256:b25e8f6c5fa1b1da422a35b1fb1510b430797dbce898277cbf7995796c437d22 to pinned sha256:3a9ac57b8868d65a89b0475be25d49e2b6fab779fe1c55af519889b2dc7064db. The fixture is the packaged evaluator pack, and this diff does not change ranking or that pin. Not re-pinned.

Skipped

memory_tests.rs is dirty in worktree /fast/tmp/fleet-graph-activation-allowance (fleet/graph-activation-allowance). Serving, registry, and reconcile files stay with PR #2452 (cursor/seal-serving-latency-669e) and PR #2596 (agent/master-unpushed-20260928). Index exclude/include files stay with PR #2514 (fleet/fix-unowned-issues-index-settings), including tracedecay-graph-query lib.rs and verified_query_source_tests.rs.

[changeset-bot]

⚠️ No Changeset found

Latest commit: f9dc9c7

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-29T09:52:33.180544Z	1f57ede	PR opened

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 1f57ede8a2

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Avoid asserting a raced activation count

On a slow or loaded test runner, the first text projection and serving swap can take longer than the test-only 50 ms activation backoff; the retry wake can then start a second activation before wait_for_seated_generation() observes the newly seated generation. The serving behavior remains correct, but the count is already 2 and this exact assertion fails nondeterministically. Synchronize observation before allowing the retry, or assert only the behavior required by this test rather than an unsynchronized attempt count.

Useful? React with 👍 / 👎.