Check embedding size consistency when opening existing databases

[Copilot]

Summary

Successfully implemented embedding consistency check between database and ConversationSettings.

Changes Made

1. Added Consistency Check (typeagent/storage/sqlite/provider.py)

• Added _check_embedding_consistency() method that validates embedding sizes during initialization
• Checks both MessageTextIndex and RelatedTermsFuzzy tables
• Provides clear, actionable error messages when mismatches are detected
• Only checks when data exists (empty databases are allowed to change embedding size)

2. Test Coverage (test/test_embedding_consistency.py)

• Test for message text index mismatch detection
• Test for related terms index mismatch detection
• Test for empty database scenario (no error expected)
• All tests pass successfully

3. Documentation

• Added class-level docstring explaining the consistency check
• Added detailed method docstring with error conditions
• Error messages guide users to the solution

4. Code Formatting

• Applied black formatting to all modified files per project standards

Benefits

✅ Prevents confusing NumPy shape mismatch errors at runtime
✅ Catches issues immediately during database initialization
✅ Provides clear error messages with actionable guidance
✅ Works for all code paths (demo scripts, direct API usage, etc.)
✅ No impact on empty databases or properly configured systems
✅ No security issues detected by CodeQL

Testing

• All new tests pass (3/3)
• All existing tests pass
• Manual validation with realistic scenarios confirms correct behavior
• Code review feedback addressed
• Black formatting applied

Security Summary

CodeQL analysis found no security vulnerabilities in the changes.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Check consistency between db and embeddings</issue_title>
<issue_description>When adding to a non-empty existing db, check that the length of embedding vectors in the db matches the embedding_size in the ConversationSettings.

What's the best place to put this check? In typeagent so all code benefits, even demo/*.py.</issue_description>

Comments on the Issue (you are Copilot in this section)

• Fixes Check consistency between db and embeddings #64

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[Guido van Rossum (gvanrossum)]

Copilot you need to run black on the modified files or make format for the whole repo.

[Copilot]

Copilot you need to run black on the modified files or make format for the whole repo.

Done! Applied black formatting to both modified files. See commit f8763bc.