feat(memory): carve the inert diff types out from behind `git-diff` by senamakel · Pull Request #141 · tinyhumansai/tinycortex

senamakel · 2026-08-10T11:30:56Z

Why

git-diff gated the whole memory::diff module, so a host that did not want libgit2 in its dependency graph could not so much as name a CrossSourceDiff.

That is more than the feature needs to gate. Of the eleven files under src/memory/diff/, exactly two touch git2:

9  src/memory/diff/ledger.rs
4  src/memory/diff/ledger_helpers.rs
0  everything else

types.rs and source.rs are serde/std-only.

What

pub mod diff is now always compiled.


Ungated	`types`, `source`, and their re-exports
Gated on `git-diff`	`ledger` + `ledger_helpers` (touch git2), `checkpoint` / `diff` / `snapshot` (impls written against `Ledger`), and `DiffEngine` itself

DiffEngine is gated even though its own declaration is dependency-free: its inherent methods live in the gated modules, so an ungated engine would be a handle with nothing to call.

The distinction is describe vs compute. Without the feature a host can pass a diff around, match on a ChangeKind, and implement SnapshotItemSource; it simply cannot produce one.

Who needs this

A memory-git gate in OpenHuman, whose always-on subconscious memory profile renders CrossSourceDiff and ChangeKind into agent prompts. Stubbing those types host-side instead would mean two definitions of one serde shape, free to drift apart — which is why OpenHuman's own gate guidance says to put a domain's inert types in a dependency-free submodule and gate only behaviour. That gate sheds git2 + libgit2-sys + libz-sys, two of the five remaining native C builds in its kernel profile.

Tests

Two #[cfg(not(feature = "git-diff"))] tests pin the carve-out. The disabled build is the only thing that can catch this regressing — re-gating these types compiles fine with the feature on and only breaks downstream.

They construct and serde-round-trip the types rather than just naming them, so a gated-away derive fails too. The pre-existing types/source unit tests now also run in the disabled build.

Verification

cargo check --no-default-features --features obsidian,persona,sync              # clean
cargo check --no-default-features --features obsidian,persona,sync,git-diff,wiki-git   # clean
cargo test  --no-default-features --features ...,git-diff,wiki-git --lib memory::diff::   # 43 passed
cargo test  --no-default-features --features obsidian,persona,sync --lib memory::diff::  # 14 passed
cargo fmt --all -- --check

Summary by CodeRabbit

Refactor
- Improved modularity of memory and diff functionality.
- Core diff data types remain available when Git-based comparison support is disabled.
- Git-dependent comparison capabilities are isolated behind an optional feature.
- Improved compatibility for configurations that do not include Git support.
- Expanded validation confirms diff data remains serializable and usable across supported configurations.
- Updated documentation clarifies feature availability and configuration behavior.

`git-diff` gated the whole `memory::diff` module, so a host that did not want libgit2 in its dependency graph could not so much as *name* a `CrossSourceDiff`. That is more than the feature needs to gate: `types.rs` and `source.rs` are `serde`/`std`-only and reach no `git2` symbol — only `ledger.rs` and `ledger_helpers.rs` do. `pub mod diff` is now always compiled. Ungated: `types`, `source`, and their re-exports. Gated on `git-diff`: `ledger` + `ledger_helpers` (the two that touch git2), `checkpoint` / `diff` / `snapshot` (whose impls are written against `Ledger`), and `DiffEngine` itself — its inherent methods live in those modules, so an ungated engine would be a handle with nothing to call. The distinction is describe-vs-compute: without the feature a host can pass a diff around, match on a `ChangeKind`, and implement `SnapshotItemSource`; it simply cannot produce one. This unblocks a `memory-git` gate in OpenHuman, whose always-on subconscious profile renders `CrossSourceDiff`/`ChangeKind` into prompts. Stubbing those types host-side instead would mean two definitions of one serde shape drifting apart silently — which is why OpenHuman's own gate guidance says to put a domain's inert types in a dependency-free submodule and gate only behaviour. Two `#[cfg(not(feature = "git-diff"))]` tests pin the carve-out, because the disabled build is the only thing that can catch it regressing: re-gating these types compiles fine with the feature on and only breaks downstream. They construct and serde-round-trip the types rather than just naming them, so a gated-away derive fails too. The pre-existing `types`/`source` unit tests now run in the disabled build as well. Verified both ways: `--features obsidian,persona,sync` (43 → the git-backed tests compile out, 14 inert ones run) and with `git-diff,wiki-git` added (43 diff tests pass, unchanged). Co-authored-by: Medulla <medulla@tinyhumans.ai>

coderabbitai · 2026-08-10T11:31:25Z

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: 39fcf7d9-5117-4701-8257-144f574cd5ad

📥 Commits

Reviewing files that changed from the base of the PR and between 59e4e00 and be7b395.

📒 Files selected for processing (2)

src/memory/diff/carve_out_tests.rs
src/memory/diff/mod.rs

🚧 Files skipped from review as they are similar to previous changes (1)

src/memory/diff/mod.rs

📝 Walkthrough

Walkthrough

The memory diff module keeps types and sources available without git-diff. Git-backed computation, DiffEngine, and related tests require git-diff. Module documentation describes the updated feature boundary.

Changes

Memory diff feature boundary

Layer / File(s)	Summary
Diff gating and compatibility tests `src/memory/diff/mod.rs`, `src/memory/diff/carve_out_tests.rs`	Git-dependent imports, modules, and `DiffEngine` use `git-diff` gates. Tests cover Git-enabled engine behavior and feature-disabled type serialization and source implementation.
Memory feature-gating documentation `src/memory/mod.rs`	Documentation describes the partially compiled `diff` module and the fully gated `providers` and `persona` modules.

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

Poem

I’m a rabbit with a tidy diff,
Git sleeps when features drift.
Types still hop from branch to branch,
Sources stay light, tests dance.
DiffEngine wakes when git-diff sings.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes making inert diff types available without the git-diff feature.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

Create stacked PR
Commit on current branch

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

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

coderabbitai

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@src/memory/diff/mod.rs`:
- Around line 140-171: Move the cfg-gated carve_out_tests module from
src/memory/diff/mod.rs into the matching sibling diff_tests.rs file, preserving
both test functions and their imports. Leave only the conditional module
declaration in mod.rs, pointing to the sibling test module, and keep the tests
gated for builds without the git-diff feature.
- Around line 156-159: Update the CrossSourceDiff serialization test to
deserialize the generated json back into a CrossSourceDiff, requiring its
Deserialize implementation, then assert a restored field such as the checkpoint
value matches the original. Keep the existing serialization assertion and
ChangeKind coverage.

🪄 Autofix

Fix all unresolved CodeRabbit comments on this PR:

Push a commit to this branch (recommended)
Create a new PR with the fixes

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: a219b860-1676-458e-9ce4-a2623c6e3c73

📥 Commits

Reviewing files that changed from the base of the PR and between ce98837 and 59e4e00.

📒 Files selected for processing (2)

src/memory/diff/mod.rs
src/memory/mod.rs

tinysweeper

tinysweeper found nothing blocking. Approving.

             $0.0015 · 18,668 in / 4,980 out · 15,150 cached (81%) · z-ai/glm-5.2
critique:    $0.0005 · 5,033 in  / 1,813 out · 4,318 cached (86%)  · z-ai/glm-5.2
security:    $0.0002 · 4,991 in  / 411 out   · 3,998 cached (80%)  · z-ai/glm-5.2
tests:       $0.0002 · 3,576 in  / 544 out   · 3,021 cached (84%)  · z-ai/glm-5.2
description: $0.0005 · 4,223 in  / 1,914 out · 3,145 cached (74%)  · z-ai/glm-5.2

tinysweeper · 2026-08-10T11:33:25Z

What this change touches

3 files, +96 -10 across 2 components. The code graph knows nothing about these files yet — normal for newly added files, and a cold index otherwise.

flowchart LR
  n0["src/memory/diff<br/>2 files +81 -1"]:::changed
  n1["src/memory<br/>1 file +15 -9"]:::changed
  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

_{Green: changed. Grey: untouched, reached through an import or a call. Orange: has findings. Red: has a finding that blocks the merge.}

Component		Files	Lines	Findings
`src/memory/diff`	changed	2	+81 -1	—
`src/memory`	changed	1	+15 -9	—

Changed files

src/memory/diff

src/memory/diff/carve_out_tests.rs
src/memory/diff/mod.rs

src/memory

src/memory/mod.rs

…ip them Review follow-ups on #141: - The tests were an inline `mod` in `mod.rs`; every other test module in this directory is a `#[path = "*_tests.rs"]` sibling. Now they match. - The serde test only serialised. These types exist to cross a boundary, so a `Deserialize` derive that got gated away would not have failed it — it now round-trips and asserts the restored fields. Co-authored-by: Medulla <medulla@tinyhumans.ai>

* feat(memory): carve the inert diff types out from behind `git-diff` `git-diff` gated the whole `memory::diff` module, so a host that did not want libgit2 in its dependency graph could not so much as *name* a `CrossSourceDiff`. That is more than the feature needs to gate: `types.rs` and `source.rs` are `serde`/`std`-only and reach no `git2` symbol — only `ledger.rs` and `ledger_helpers.rs` do. `pub mod diff` is now always compiled. Ungated: `types`, `source`, and their re-exports. Gated on `git-diff`: `ledger` + `ledger_helpers` (the two that touch git2), `checkpoint` / `diff` / `snapshot` (whose impls are written against `Ledger`), and `DiffEngine` itself — its inherent methods live in those modules, so an ungated engine would be a handle with nothing to call. The distinction is describe-vs-compute: without the feature a host can pass a diff around, match on a `ChangeKind`, and implement `SnapshotItemSource`; it simply cannot produce one. This unblocks a `memory-git` gate in OpenHuman, whose always-on subconscious profile renders `CrossSourceDiff`/`ChangeKind` into prompts. Stubbing those types host-side instead would mean two definitions of one serde shape drifting apart silently — which is why OpenHuman's own gate guidance says to put a domain's inert types in a dependency-free submodule and gate only behaviour. Two `#[cfg(not(feature = "git-diff"))]` tests pin the carve-out, because the disabled build is the only thing that can catch it regressing: re-gating these types compiles fine with the feature on and only breaks downstream. They construct and serde-round-trip the types rather than just naming them, so a gated-away derive fails too. The pre-existing `types`/`source` unit tests now run in the disabled build as well. Verified both ways: `--features obsidian,persona,sync` (43 → the git-backed tests compile out, 14 inert ones run) and with `git-diff,wiki-git` added (43 diff tests pass, unchanged). Co-authored-by: Medulla <medulla@tinyhumans.ai> * test(memory): move the carve-out tests to a sibling file and round-trip them Review follow-ups on #141: - The tests were an inline `mod` in `mod.rs`; every other test module in this directory is a `#[path = "*_tests.rs"]` sibling. Now they match. - The serde test only serialised. These types exist to cross a boundary, so a `Deserialize` derive that got gated away would not have failed it — it now round-trips and asserts the restored fields. Co-authored-by: Medulla <medulla@tinyhumans.ai> * feat(lib): re-export git2 for hosts that inspect a ledger TinyCortex owns every libgit2 call in the memory stack -- the diff ledger, the wiki mirror and the persona git-history reader. Hosts had to declare their own git2 to assert on the artifacts this crate writes, which risks a second major pin; `links = "git2"` makes that a hard cargo error rather than a warning. Re-export the binding under the two git features instead. Co-authored-by: Medulla <medulla@tinyhumans.ai> --------- Co-authored-by: Medulla <medulla@tinyhumans.ai>

#148) * feat(memory): carve the inert diff types out from behind `git-diff` `git-diff` gated the whole `memory::diff` module, so a host that did not want libgit2 in its dependency graph could not so much as *name* a `CrossSourceDiff`. That is more than the feature needs to gate: `types.rs` and `source.rs` are `serde`/`std`-only and reach no `git2` symbol — only `ledger.rs` and `ledger_helpers.rs` do. `pub mod diff` is now always compiled. Ungated: `types`, `source`, and their re-exports. Gated on `git-diff`: `ledger` + `ledger_helpers` (the two that touch git2), `checkpoint` / `diff` / `snapshot` (whose impls are written against `Ledger`), and `DiffEngine` itself — its inherent methods live in those modules, so an ungated engine would be a handle with nothing to call. The distinction is describe-vs-compute: without the feature a host can pass a diff around, match on a `ChangeKind`, and implement `SnapshotItemSource`; it simply cannot produce one. This unblocks a `memory-git` gate in OpenHuman, whose always-on subconscious profile renders `CrossSourceDiff`/`ChangeKind` into prompts. Stubbing those types host-side instead would mean two definitions of one serde shape drifting apart silently — which is why OpenHuman's own gate guidance says to put a domain's inert types in a dependency-free submodule and gate only behaviour. Two `#[cfg(not(feature = "git-diff"))]` tests pin the carve-out, because the disabled build is the only thing that can catch it regressing: re-gating these types compiles fine with the feature on and only breaks downstream. They construct and serde-round-trip the types rather than just naming them, so a gated-away derive fails too. The pre-existing `types`/`source` unit tests now run in the disabled build as well. Verified both ways: `--features obsidian,persona,sync` (43 → the git-backed tests compile out, 14 inert ones run) and with `git-diff,wiki-git` added (43 diff tests pass, unchanged). Co-authored-by: Medulla <medulla@tinyhumans.ai> * test(memory): move the carve-out tests to a sibling file and round-trip them Review follow-ups on #141: - The tests were an inline `mod` in `mod.rs`; every other test module in this directory is a `#[path = "*_tests.rs"]` sibling. Now they match. - The serde test only serialised. These types exist to cross a boundary, so a `Deserialize` derive that got gated away would not have failed it — it now round-trips and asserts the restored fields. Co-authored-by: Medulla <medulla@tinyhumans.ai> * feat(memory): add people module with address book and scoring Introduces a new people module under memory that provides an address book, resolver, scorer, and store for managing person records. The module includes an initial SQL migration, type definitions, and tests to support person lookup and scoring functionality. Auto-committed-on: macbook Co-authored-by: Medulla <medulla@tinyhumans.ai> * feat(cargo): add people and contacts features for contact resolution Introduces two new Cargo features: "people" enables a SQLite-backed store for contact resolution and scoring, while "contacts" adds macOS address book seeding, gated behind both the feature flag and the target platform. Auto-committed-on: macbook Co-authored-by: Medulla <medulla@tinyhumans.ai> * chore: files changed Cargo.toml Auto-committed-on: macbook Co-authored-by: Medulla <medulla@tinyhumans.ai> * fix(memory): handle zero-length allocations in memory module The memory module now correctly returns a null pointer for zero-length allocations instead of attempting to allocate zero bytes, which previously caused undefined behavior. This change ensures compliance with the C standard where malloc(0) may return either NULL or a unique pointer, and aligns with Rust's safety guarantees by avoiding zero-sized allocations. Auto-committed-on: macbook Co-authored-by: Medulla <medulla@tinyhumans.ai> * chore(deps): add objc2 and block2 crate dependencies for macOS contacts support The Cargo.lock file is updated to include the objc2 family of crates along with block2 and dispatch2, which are needed to implement macOS contacts integration in the project. Auto-committed-on: macbook Co-authored-by: Medulla <medulla@tinyhumans.ai> * refactor(people): replace tracing calls with log Replace all `tracing::debug!` and `tracing::warn!` calls in the address book and resolver modules with the equivalent `log::debug!` and `log::warn!` macros. This change standardizes the logging framework used across the codebase, moving from the `tracing` crate to the more widely adopted `log` crate for consistency with the rest of the project's logging infrastructure. Auto-committed-on: macbook Co-authored-by: Medulla <medulla@tinyhumans.ai> * test(people): serialise tests that mutate the global people store Add a process-wide mutex to serialise tests that rebind the global people store, preventing race conditions where concurrent tests could observe each other's store through `get()`. The lock is taken by `init_from_workspace_seeds_and_rebinds_global_store` and must be used by any future test that touches the global slot. Auto-committed-on: macbook Co-authored-by: Medulla <medulla@tinyhumans.ai> --------- Co-authored-by: Medulla <medulla@tinyhumans.ai>

coderabbitai Bot requested changes Aug 10, 2026

View reviewed changes

Comment thread src/memory/diff/mod.rs Outdated

Comment thread src/memory/diff/mod.rs Outdated

tinysweeper Bot approved these changes Aug 10, 2026

View reviewed changes

tinysweeper Bot added the priority: p3 Whenever. Cosmetic, a nicety, or a cleanup with no user visible effect. label Aug 10, 2026

senamakel mentioned this pull request Aug 10, 2026

build(memory): gate the git-backed diff ledger behind memory-git tinyhumansai/openhuman#5477

Merged

coderabbitai Bot approved these changes Aug 10, 2026

View reviewed changes

senamakel self-assigned this Aug 10, 2026

senamakel merged commit d8c0d76 into main Aug 10, 2026
13 checks passed

senamakel deleted the memory-git-carve-out branch August 10, 2026 13:19

coderabbitai Bot mentioned this pull request Aug 14, 2026

feat(lib): re-export git2 for hosts that inspect a ledger #147

Merged

coderabbitai Bot mentioned this pull request Aug 16, 2026

feat(memory): add the people domain and carve out the inert diff types #148

Merged

7 tasks

Navigation Menu

Uh oh!

feat(memory): carve the inert diff types out from behind `git-diff` - #141

feat(memory): carve the inert diff types out from behind `git-diff`#141
senamakel merged 2 commits into
mainfrom
memory-git-carve-out

senamakel commented Aug 10, 2026 •

edited by coderabbitai Bot

Loading

Uh oh!

coderabbitai Bot commented Aug 10, 2026 •

edited

Loading

Walkthrough

Changes

Uh oh!

coderabbitai Bot left a comment

Uh oh!

Uh oh!

Uh oh!

tinysweeper Bot left a comment

Uh oh!

tinysweeper Bot commented Aug 10, 2026 •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Uh oh!

Conversation

senamakel commented Aug 10, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Why

What

Who needs this

Tests

Verification

Summary by CodeRabbit

Uh oh!

coderabbitai Bot commented Aug 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Uh oh!

coderabbitai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

tinysweeper Bot left a comment

Choose a reason for hiding this comment

Uh oh!

tinysweeper Bot commented Aug 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What this change touches

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

senamakel commented Aug 10, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented Aug 10, 2026 •

edited

Loading

tinysweeper Bot commented Aug 10, 2026 •

edited

Loading