Embedded engine: off-executor SQL, KV read/write symmetry, KV conformance (#75)

[YellowSnnowmann]

Summary

The embedded-engine batch of #75 — both machine-verified findings plus the coverage gap that hid one of them.

• Off-executor SQL: every Memory-trait method on UnifiedMemory that ran synchronous rusqlite inline now hops to spawn_blocking, copying the profile/episodic families' exact idiom (extracted sync bodies, owned params, contexted joins). recall's two synchronous fts5 fetches hop the same way; forget is split (sync lookup wrapped, internally-awaiting delete_document untouched). The repo documented this offloading in three places while the hot path did the opposite — the claims are now true, and the two Cargo.toml comments are narrowed to the families that actually hop.
• KV symmetry: kv_get/kv_list canonicalize the caller key/prefix through the same canonical_identifier the write-path shim applies — put→get and put→list stop missing rewritten keys while kv_delete (always shim-routed) worked, the incoherent split Remote + embedded audit at v1.1.0: Supermemory re-store 400s (PATCH drops containerTag), keyed-seam gaps, and verified adapter defects #75 flagged. Prefix matching documented as over canonical stored keys.
• KV conformance: assert_kv_round_trip joins the suite (put → get → list-by-prefix → delete → gone), one leg keyed by a formatted national-ID shape the canonicalizer provably rewrites, guarded by assert_ne! so predicate drift fails loudly. The KV family had zero conformance coverage — how the asymmetry survived. Adapter tests pin symmetry under a rewritten namespace too.

Deliberate scope choices: the spec's Luhn digit-run fixture would have been vacuous (bare digit runs are deliberately outside the strict PII gate — the #1201 Luhn-FP history), so the fixture is a shape the gate really rewrites, machine-checked. A keyed KV record fetch (to stop kv_get's namespace listing) needs updated_at from the vendored tinycortex's keyed getters, which don't return it — deferred until a tinycortex-side additive change exists.

Public API and behavior changes

Additive only: assert_kv_round_trip is a new public conformance case (wired into assert_provider, capability-gated on as_graph()). Behavior: embedded Memory calls no longer block executor threads; KV reads find canonicalizer-rewritten keys (previously permanently unreadable).

Validation

• cargo fmt --all -- --check — clean
• cargo clippy --all-targets --all-features -- -D warnings — zero
• cargo build --all-targets --all-features — clean
• cargo test --all-features — green; red-proof: reverting the engine fix reds 3 tests (both adapter KV tests + the suite case through the full provider)

Companion PRs: #76 (remote adapter batch, same issue). Closes the embedded-batch checkboxes of #75.

🤖 Generated with Claude Code

https://claude.ai/code/session_01TVSoZqKtJzAqTnYk4WBuEV

[coderabbitai]

Warning

Review limit reached

Your included review limit has been reached.

You’re in a promotional period — use the checkbox below to run this review for free:

• Run review for free

_{On-demand reviews are free for the next 30 days. After that, they cost $0.25 per reviewed file.}

How can I continue?

Run this review now using the option above, or comment @coderabbitai review --use-credits.

You can also wait for the limit to reset (next review available in 42 minutes), then comment @coderabbitai review or push new commits to the PR.

An organization admin can change what happens after included review limits in Billing.

How do review limits work?

CodeRabbit enforces per-developer PR review limits within each organization.

For paid Pro and Pro+ reviews, CodeRabbit uses a developer's included PR review attempts over the past 7 days to set the current hourly allowance. At typical activity levels, the full plan allowance applies. Higher sustained activity can lower the allowance until earlier attempts leave the 7-day window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 720372ad-0a27-447d-aa05-5f69515272fa

📥 Commits

Reviewing files that changed from the base of the PR and between 1d501fb and 8652870.

📒 Files selected for processing (6)

• crates/tinymemory-conformance/src/lib.rs
• crates/tinymemory-conformance/src/suite/mod.rs
• crates/tinymemory-core/src/store/memory_trait.rs
• crates/tinymemory-tinycortex/Cargo.toml
• crates/tinymemory-tinycortex/src/engine/mod.rs
• crates/tinymemory-tinycortex/tests/full_provider_conformance.rs

---

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 · 718 embedded · openrouter/openai/text-embedding-3-small

[tinysweeper]

How this change flows

2 changed behaviours across 10 relationships. 6 surrounding behaviours are shown (60 graph nodes walked). 42 further behaviours left out to keep the diagram readable.

flowchart LR
  n0["assert_awkward_content_round_trips<br/>changed"]:::changed
  n1["assert_provider<br/>changed"]:::changed
  n2["fresh_mem"]:::impacted
  n3["ns"]:::impacted
  n4["...es_external_sync_taint_for_origin_upgrade"]:::impacted
  n5["iter"]:::impacted
  n6["cleanup"]:::impacted
  n7["recall"]:::impacted
  n0 -->|calls| n3
  n0 -->|calls| n5
  n0 -->|calls| n6
  n1 -->|calls| n0
  n4 -->|calls| n2
  n4 -->|tests| n2
  n4 -->|calls| n5
  n4 -->|tests| n5
  n4 -->|calls| n7
  n4 -->|tests| n7
  classDef changed fill:#0d4429,stroke:#238636,color:#e6edf3
  classDef impacted fill:#161b22,stroke:#6e7681,color:#c9d1d9
  classDef flagged fill:#5a1e02,stroke:#d93f0b,color:#ffffff
  classDef blocking fill:#67060c,stroke:#f85149,color:#ffffff

Loading

_{Green: changed behaviour. Grey: surrounding behaviour. Arrows name the call, use, implementation, or test relationship. Orange: has findings. Red: has a finding that blocks the merge.}