fix(memory): align stale capability tests with the pinned-artifact fix

[senamakel]

Summary

• Fixes three tests that still hardcoded the pre-fix(memory): advertise what the pinned module serves, not the whole contract #5620 "module driver advertises the whole 18-family contract" expectation, which has been red on main since PR fix(memory): advertise what the pinned module serves, not the whole contract #5620 merged.
• No production code changes — modules::memory::ARTIFACT_CAPABILITIES is untouched.
• Unblocks Rust Core Coverage (cargo-llvm-cov) / PR CI Gate on main and every PR based on it.

Problem

PR #5620 ("fix(memory): advertise what the pinned module serves, not the whole contract", merged 2026-08-20T12:41Z) corrected ModuleMemoryProvider::capabilities() to advertise only the 13 families the pinned tinymemory v1.0.1 artifact actually serves (per issue #5598), instead of falsely claiming the full 18-family contract (Capabilities::all()). It updated memory_tests.rs to match the corrected behavior, but three other test files independently hardcoded the old, now-incorrect full-contract expectation and were not updated in that PR:

• core::cli_capability::tests::bound_driver_probe_reports_the_default_module_driver — asserted binding.capabilities() == Capabilities::all()
• openhuman::memory::binding::tests::module_binding_advertises_every_family — asserted every Capability::ALL variant is advertised and advertised == unbound_default_capabilities()
• openhuman::memory::ops::provider::tests::bound_driver_status_reports_id_class_contract_and_capabilities — asserted the wire status lists all 18 capability strings

All three have failed on every main run since #5620 merged (confirmed against main run 32373660712, the merge of #5621: identical 12116 passed; 3 failed; 106 ignored line, same three tests).

Solution

Update all three tests to assert the corrected 13-family set instead of the full 18-family contract, mirroring the reasoning #5620 already established and got reviewed/merged: the pinned v1.0.1 artifact has no bus member for people, chunks, retrieval, profile, or episodic (#5598), so claiming them is the bug, not the fix. Where it strengthens the test, I also assert the 5 not-yet-served families are explicitly absent, so a future artifact bump that adds them is what should widen this — deliberately, per the comment already in modules/memory.rs.

I did not touch ARTIFACT_CAPABILITIES or any other production code — this is strictly bringing stale test assertions in line with already-accepted, already-merged behavior.

Submission Checklist

• Tests added or updated (happy path + at least one failure / edge case) — three existing tests corrected; each now also asserts the negative case (the 5 not-yet-served families are absent), which is new coverage exercising Staging: tinymemory capability mismatch (8191 vs 262143) — memory_tree, memory_store_raw_chunks, memory_diff all failing #5598's guard from a second angle.
• N/A: Diff coverage ≥ 80% — test-only change to already-covered lines; no new production lines to cover.
• N/A: Coverage matrix updated — behaviour-only test correction, no feature added/removed/renamed.
• N/A: All affected feature IDs listed under ## Related — no feature IDs affected.
• No new external network dependencies introduced.
• N/A: Manual smoke checklist — no release-cut surface touched.
• Linked issue — see ## Related (does not close Staging: tinymemory capability mismatch (8191 vs 262143) — memory_tree, memory_store_raw_chunks, memory_diff all failing #5598, which tracks the longer-term artifact-capability gap; this only fixes the stale test debt from fix(memory): advertise what the pinned module serves, not the whole contract #5620).

Impact

• CI-only: unblocks Rust Core Coverage (cargo-llvm-cov) on main (and therefore PR CI Gate on every open PR based on it). No runtime/platform behavior change.

Related

• Follow-up PR(s)/TODOs: none. Issue Staging: tinymemory capability mismatch (8191 vs 262143) — memory_tree, memory_store_raw_chunks, memory_diff all failing #5598 (re-pinning tinymemory to a release that serves all 18 families) remains open and tracked separately; it is not blocking and this PR does not attempt to resolve it.

---

AI Authored PR Metadata (required for Codex/Linear PRs)

Linear Issue

• Key: N/A
• URL: N/A

Commit & Branch

• Branch: fix/stale-artifact-capability-tests
• Commit SHA: e546845

Validation Run

• pnpm --filter openhuman-app format:check — N/A: no app/ files touched
• pnpm typecheck — N/A: no TypeScript touched
• Focused tests: cargo test --lib -F modules -- bound_driver_probe_reports_the_default_module_driver module_binding_advertises_the_pinned_artifacts_families bound_driver_status_reports_id_class_contract_and_capabilities — 3 passed, 0 failed
• Rust fmt/check (if changed): cargo fmt --check clean on all three touched files
• N/A: Tauri fmt/check (if changed) — no Tauri shell files touched

Validation Blocked

• command: N/A
• error: N/A
• impact: N/A

Behavior Changes

• Intended behavior change: none — test assertions only, correcting stale expectations to match already-merged production behavior from fix(memory): advertise what the pinned module serves, not the whole contract #5620.
• User-visible effect: none.

Summary by CodeRabbit

• Tests
  • Updated capability validation to reflect the capabilities currently advertised by supported memory artifacts.
  • Added checks ensuring required capabilities are present and unsupported capabilities are excluded.
  • Confirmed advertised capabilities remain within the defined contract.
  • Preserved full capability behavior for unbound contexts.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 9520c5c9-1bff-44bf-8348-782df8e6b23e

📥 Commits

Reviewing files that changed from the base of the PR and between 2b220b8 and e546845.

📒 Files selected for processing (3)

• src/core/cli_capability_tests.rs
• src/openhuman/memory/binding_tests.rs
• src/openhuman/memory/ops/provider.rs

Included review availability: Your plan provides up to 10 included reviews per hour; 9 remain after this review.

---

📝 Walkthrough

Walkthrough

Capability tests now match the pinned TinyMemory artifact. They validate 13 advertised capability families, reject five unsupported families, enforce contract bounds, and preserve the full capability set for unbound contexts.

Changes

TinyMemory capability validation

Layer / File(s)	Summary
Capability boundary validation src/core/cli_capability_tests.rs, src/openhuman/memory/binding_tests.rs, src/openhuman/memory/ops/provider.rs	Tests now expect the artifact’s 13 supported capability families, reject chunks, episodic, people, profile, and retrieval, and verify contract bounds. Unbound contexts still expose all capabilities.

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: ⚪ Minimal · up to e5468

This PR only aligns existing tests with the already-deployed 13-family memory capability behavior and does not change runtime behavior; no actionable merge-blocking risk remains after normal checks and review.

Suggested reviewers: al629176

Poem

I’m a rabbit with a tidy list,
Thirteen capabilities now exist.
Five stay out beyond the gate,
Contract bounds keep the sets in shape.
The TinyMemory tests agree—
No extra bits for me! 🐇

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Linked Issues check	⚠️ Warning	The tests reflect the pinned artifact, but the linked issue requires re-pinning the artifact or resolving the capability mismatch.	Re-pin tinymemory to an artifact that advertises all 18 required capability families, or link this PR to a test-only issue.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly describes updating stale memory capability tests to match the pinned artifact.
Out of Scope Changes check	✅ Passed	All changes update memory capability tests and remain within the stated test-alignment objective.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 4 functions across 3 files.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[tinysweeper]

tinysweeper found nothing blocking. Approving.

$0.0000 · 0 in / 0 out · 272 embedded · openrouter/openai/text-embedding-3-small