Add content part support in prep for multi-modal embedding models

[chrisboulton]

The shape of this matches media for SeqAppend - from the modelsocket perspective it's something like:

{
  "request": "seq_command",
  "cid": "embed-1",
  "seq_id": "seq-123",
  "data": {
    "command": "embed",
    "inputs": [
      {
        "content": [
          {
            "text": "Find products similar to this bicycle"
          },
          {
            "media": {
              "uri": "https://example.com/bicycle.png",
              "mime_type": "image/png",
              "detail": "high"
            }
          }
        ]
      }
    ],
    "input_type": "image",
    "dimensions": 1024,
    "normalize": true,
    "truncate": false
  }
}

Summary by CodeRabbit

• New Features

  • Added support for embedding content that includes text and media.
  • Media inputs can now reference a URI or embedded content, with optional metadata such as MIME type, detail, and hash.
  • Added reliable serialization and deserialization for embedding requests.

• Bug Fixes

  • Improved consistency in handling media inputs across supported operations.

• Chores

  • Updated the release version to 0.5.1.
  • Added macOS system files to the ignored files list.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 7b49f2f0-7513-4f7f-9bba-a6b565ea59cf

📥 Commits

Reviewing files that changed from the base of the PR and between cf831ce and b6f5b08.

⛔ Files ignored due to path filters (1)

• Cargo.lock is excluded by !**/*.lock

📒 Files selected for processing (3)

• .gitignore
• Cargo.toml
• src/protocol.rs

---

📝 Walkthrough

Walkthrough

The protocol now shares media input modeling between append and embed requests, supports ordered text/media embedding content, and adds round-trip tests for URI and blob media. The crate version and repository ignore rules are also updated.

Changes

Embedding protocol support

Layer / File(s)	Summary
Media and embedding content contract src/protocol.rs	A shared MediaInput type backs append and embed media aliases, while EmbeddingContentPart supports ordered text and media content.
Embedding serialization validation src/protocol.rs	Round-trip tests cover simple embedding inputs and content containing URI- or blob-backed media.
Package and repository metadata Cargo.toml, .gitignore	The crate version changes to 0.5.1, and .DS_Store is ignored alongside .venv.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Possibly related PRs

• mixlayer/modelsocket-rs#23: Both changes update media attachment modeling and serialization in src/protocol.rs.

Suggested reviewers: zackangelo

🚥 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 summarizes the main change: adding content part support for multimodal embedding models.
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

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch mm-embed

---

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