<!-- .github/pull_request_template.md -->
## Description
<!--
Please provide a clear, human-generated description of the changes in
this PR.
DO NOT use AI-generated descriptions. We want to understand your thought
process and reasoning.
-->
## Acceptance Criteria
<!--
* Key requirements to the new feature or modification;
* Proof that the changes work and meet the requirements;
* Include instructions on how to verify the changes. Describe how to
test it locally;
* Proof that it's sufficiently tested.
-->
## Type of Change
<!-- Please check the relevant option -->
- [ ] Bug fix (non-breaking change that fixes an issue)
- [ ] New feature (non-breaking change that adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to change)
- [ ] Documentation update
- [ ] Code refactoring
- [ ] Performance improvement
- [ ] Other (please specify):
## Screenshots/Videos (if applicable)
<!-- Add screenshots or videos to help explain your changes -->
## Pre-submission Checklist
<!-- Please check all boxes that apply before submitting your PR -->
- [ ] **I have tested my changes thoroughly before submitting this PR**
- [ ] **This PR contains minimal changes necessary to address the
issue/feature**
- [ ] My code follows the project's coding standards and style
guidelines
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] I have added necessary documentation (if applicable)
- [ ] All new and existing tests pass
- [ ] I have searched existing PRs to ensure this change hasn't been
submitted already
- [ ] I have linked any relevant issues in the description
- [ ] My commits have clear and descriptive messages
## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added configurable chunks-per-batch to control per-batch processing
size via CLI flag, API payload, and configuration; defaults are now
driven by config with an automatic fallback.
* **Style / Documentation**
* Updated contribution/style guidelines (formatting, line length,
string-quote rule, pre-commit note).
* **Tests**
* Updated CLI tests to verify propagation of the new chunks-per-batch
parameter.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- .github/pull_request_template.md -->
## Description
- Add delete ability for Neo4j Aura
- Refactor Neo4j Aura to use aiohttp to make async requests and perform
better
## Acceptance Criteria
<!--
* Key requirements to the new feature or modification;
* Proof that the changes work and meet the requirements;
* Include instructions on how to verify the changes. Describe how to
test it locally;
* Proof that it's sufficiently tested.
-->
## Type of Change
<!-- Please check the relevant option -->
- [ ] Bug fix (non-breaking change that fixes an issue)
- [ ] New feature (non-breaking change that adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to change)
- [ ] Documentation update
- [ ] Code refactoring
- [ ] Performance improvement
- [ ] Other (please specify):
## Screenshots/Videos (if applicable)
<!-- Add screenshots or videos to help explain your changes -->
## Pre-submission Checklist
<!-- Please check all boxes that apply before submitting your PR -->
- [ ] **I have tested my changes thoroughly before submitting this PR**
- [ ] **This PR contains minimal changes necessary to address the
issue/feature**
- [ ] My code follows the project's coding standards and style
guidelines
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] I have added necessary documentation (if applicable)
- [ ] All new and existing tests pass
- [ ] I have searched existing PRs to ensure this change hasn't been
submitted already
- [ ] I have linked any relevant issues in the description
- [ ] My commits have clear and descriptive messages
## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Performance Improvements**
* Neo4j Aura database operations are now asynchronous, eliminating
blocking requests and improving system responsiveness during dataset
management.
* Token retrieval and database provisioning workflows now use
non-blocking asynchronous calls.
* Enhanced error handling for database API interactions.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- .github/pull_request_template.md -->
## Description
<!--
Please provide a clear, human-generated description of the changes in
this PR.
DO NOT use AI-generated descriptions. We want to understand your thought
process and reasoning.
-->
- Adds batch search support to `brute_force_triplet_search` with a new
`query_batch` parameter that accepts a list of queries in addition to
the existing single `query` parameter.
- Introduces a new `NodeEdgeVectorSearch` class that encapsulates vector
search operations, handling embedding and distance retrieval for both
single-query and batch-query modes.
- Returns `List[List[Edge]]` (one list per query) when using
`query_batch`, instead of the single `List[Edge]` format used for single
queries.
- Adds comprehensive test coverage including new test files and cases
for the `NodeEdgeVectorSearch` class, batch search functionality, and
edge cases for both single and batch modes.
- Refactors code by extracting vector search logic into the new class
and adding a helper function `_get_top_triplet_importances` to reduce
code duplication and improve maintainability.
## Acceptance Criteria
<!--
* Key requirements to the new feature or modification;
* Proof that the changes work and meet the requirements;
* Include instructions on how to verify the changes. Describe how to
test it locally;
* Proof that it's sufficiently tested.
-->
## Type of Change
<!-- Please check the relevant option -->
- [ ] Bug fix (non-breaking change that fixes an issue)
- [x] New feature (non-breaking change that adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to change)
- [ ] Documentation update
- [ ] Code refactoring
- [ ] Performance improvement
- [ ] Other (please specify):
## Screenshots/Videos (if applicable)
<!-- Add screenshots or videos to help explain your changes -->
## Pre-submission Checklist
<!-- Please check all boxes that apply before submitting your PR -->
- [x] **I have tested my changes thoroughly before submitting this PR**
- [x] **This PR contains minimal changes necessary to address the
issue/feature**
- [x] My code follows the project's coding standards and style
guidelines
- [x] I have added tests that prove my fix is effective or that my
feature works
- [ ] I have added necessary documentation (if applicable)
- [x] All new and existing tests pass
- [x] I have searched existing PRs to ensure this change hasn't been
submitted already
- [x] I have linked any relevant issues in the description
- [x] My commits have clear and descriptive messages
## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Added batch-query support to triplet search; batch returns per-query
nested results while single-query remains flat.
* Introduced a unified vector search controller to embed queries and
retrieve node/edge distances across collections.
* **Bug Fixes**
* Improved input validation and safer error handling for missing
collections and batch failures.
* Stopped adding duplicate skeleton edge links after edge creation.
* **Tests**
* Added comprehensive unit and integration tests covering single/batch
flows and edge cases.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
- Remove use_combined_context parameter from search functions
- Remove CombinedSearchResult class from types module
- Update API routers to remove combined search support
- Remove prepare_combined_context helper function
- Update tutorial notebook to remove use_combined_context usage
- Simplify search return types to always return List[SearchResult]
This removes the combined search feature which aggregated results across
multiple datasets into a single response. Users can still search across
multiple datasets and get results per dataset.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
<!-- .github/pull_request_template.md -->
## Description
<!--
Please provide a clear, human-generated description of the changes in
this PR.
DO NOT use AI-generated descriptions. We want to understand your thought
process and reasoning.
-->
## Acceptance Criteria
<!--
* Key requirements to the new feature or modification;
* Proof that the changes work and meet the requirements;
* Include instructions on how to verify the changes. Describe how to
test it locally;
* Proof that it's sufficiently tested.
-->
## Type of Change
<!-- Please check the relevant option -->
- [ ] Bug fix (non-breaking change that fixes an issue)
- [ ] New feature (non-breaking change that adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to change)
- [ ] Documentation update
- [ ] Code refactoring
- [ ] Performance improvement
- [ ] Other (please specify):
## Screenshots/Videos (if applicable)
<!-- Add screenshots or videos to help explain your changes -->
## Pre-submission Checklist
<!-- Please check all boxes that apply before submitting your PR -->
- [ ] **I have tested my changes thoroughly before submitting this PR**
- [ ] **This PR contains minimal changes necessary to address the
issue/feature**
- [ ] My code follows the project's coding standards and style
guidelines
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] I have added necessary documentation (if applicable)
- [ ] All new and existing tests pass
- [ ] I have searched existing PRs to ensure this change hasn't been
submitted already
- [ ] I have linked any relevant issues in the description
- [ ] My commits have clear and descriptive messages
## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Breaking Changes**
* Search API response simplified: combined-context result type removed
and the legacy combined-context request flag eliminated, changing
response shapes.
* **New Features**
* dataset_name added to each search result for clearer attribution.
* **Refactor**
* Search logic and return shapes streamlined for access-control and
per-dataset flows; telemetry and request parameters aligned.
* **Tests**
* Combined-context related tests removed or updated to reflect
simplified behavior.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
<!-- .github/pull_request_template.md -->
## Description
Resolve conflict and merge commits from main to dev
## Acceptance Criteria
<!--
* Key requirements to the new feature or modification;
* Proof that the changes work and meet the requirements;
* Include instructions on how to verify the changes. Describe how to
test it locally;
* Proof that it's sufficiently tested.
-->
## Type of Change
<!-- Please check the relevant option -->
- [ ] Bug fix (non-breaking change that fixes an issue)
- [ ] New feature (non-breaking change that adds functionality)
- [ ] Breaking change (fix or feature that would cause existing
functionality to change)
- [ ] Documentation update
- [ ] Code refactoring
- [ ] Performance improvement
- [ ] Other (please specify):
## Screenshots/Videos (if applicable)
<!-- Add screenshots or videos to help explain your changes -->
## Pre-submission Checklist
<!-- Please check all boxes that apply before submitting your PR -->
- [ ] **I have tested my changes thoroughly before submitting this PR**
- [ ] **This PR contains minimal changes necessary to address the
issue/feature**
- [ ] My code follows the project's coding standards and style
guidelines
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] I have added necessary documentation (if applicable)
- [ ] All new and existing tests pass
- [ ] I have searched existing PRs to ensure this change hasn't been
submitted already
- [ ] I have linked any relevant issues in the description
- [ ] My commits have clear and descriptive messages
## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **New Features**
* Add top_k to control number of search results
* Add verbose option to include/exclude detailed graphs in search output
* **Improvements**
* Examples now use pretty-printed output for clearer readability
* Startup handles migration failures more gracefully with a fallback
initialization path
* **Documentation**
* Updated contributing guidance and added explicit run instructions for
examples
* **Chores**
* Project version bumped to 0.5.1
* Adjusted frontend framework version constraint
* **Tests**
* Updated tests to exercise verbose search behavior
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
## Description
This PR adds usage frequency tracking to help identify which graph
elements (nodes) are most frequently accessed during user searches.
**Related Issue:** Closes [#1458]
**The Problem:**
When users search repeatedly, we had no way to track which pieces of
information were being referenced most often. This made it impossible
to:
- Prioritize popular content in search results
- Understand which topics users care about most
- Improve retrieval by boosting frequently-used nodes
**The Solution:**
I've implemented a system that tracks usage patterns by:
1. Leveraging the existing `save_interaction=True` flag in
`cognee.search()` which creates `CogneeUserInteraction` nodes
2. Following the `used_graph_element_to_answer` edges to see which graph
elements each search referenced
3. Counting how many times each element was accessed within a
configurable time window (default: 7 days)
4. Writing a `frequency_weight` property back to frequently-accessed
nodes
This gives us a simple numeric weight on nodes that reflects real usage
patterns, which can be used to improve search ranking, analytics
dashboards, or identifying trending topics.
**Key Design Decisions:**
- Time-windowed counting (not cumulative) - focuses on recent usage
patterns
- Configurable minimum threshold - filters out noise from rarely
accessed nodes
- Neo4j-first implementation using Cypher queries - works with our
primary production database
- Documented Kuzu limitation - requires schema changes, leaving for
future work as acceptable per team discussion
The implementation follows existing patterns in Cognee's memify pipeline
and can be run as a scheduled task or on-demand.
**Known Limitations:**
**Kuzu adapter not currently supported** - Kuzu requires properties to
be defined in the schema at node creation time, so dynamic property
updates don't work. I'm opening a separate issue to track Kuzu support,
which will require schema modifications in the Kuzu adapter. For now,
this feature works with Neo4j (our primary production database).
**Follow-up Issue:** #1993
## Acceptance Criteria
**Core Functionality:**
- ✅ `extract_usage_frequency()` correctly counts node access frequencies
from interaction data
- ✅ `add_frequency_weights()` writes `frequency_weight` property to
Neo4j nodes
- ✅ Time window filtering works (only counts recent interactions)
- ✅ Minimum threshold filtering works (excludes rarely-used nodes)
- ✅ Element type distribution tracked for analytics
- ✅ Gracefully handles unsupported adapters (logs warning, doesn't
crash)
**Testing Verification:**
1. Run the end-to-end example with Neo4j:
```bash
# Update .env for Neo4j
GRAPH_DATABASE_PROVIDER=neo4j
GRAPH_DATASET_HANDLER=neo4j_aura_dev
python extract_usage_frequency_examplepy
```
Should show frequencies extracted and applied to nodes
2. Verify in Neo4j Browser (http://localhost:7474):
```cypher
MATCH (n) WHERE n.frequency_weight IS NOT NULL
RETURN n.frequency_weight, labels(n), n.text
ORDER BY n.frequency_weight DESC LIMIT 10
```
Should return nodes with frequency weights
3. Run unit tests:
```bash
python test_usage_frequency.py
```
All tests pass (tests are adapter-agnostic and test core logic)
4. Test graceful handling with unsupported adapter:
```bash
# Update .env for Kuzu
GRAPH_DATABASE_PROVIDER=kuzu
GRAPH_DATASET_HANDLER=kuzu
python extract_usage_frequency_example.py
```
Should log warning about Kuzu not being supported but not crash
**Files Added:**
- `cognee/tasks/memify/extract_usage_frequency.py` - Core implementation
(215 lines)
- `extract_usage_frequency_example.py` - Complete working example with
documentation
- `test_usage_frequency.py` - Unit tests for core logic
- Test utilities and Neo4j setup scripts for local development
**Tested With:**
- Neo4j 5.x (primary target, fully working)
- Kuzu (gracefully skips with warning)
- Python 3.10, 3.11
- Existing Cognee interaction tracking (save_interaction=True)
**What This Solves:**
This directly addresses the need for usage-based ranking mentioned in
[#1458]. Now teams can:
- See which information gets referenced most in their knowledge base
- Build analytics dashboards showing popular topics
- Weight search results by actual usage patterns
- Identify content that needs improvement (low frequency despite high
relevance)
## Type of Change
- [x] New feature (non-breaking change that adds functionality)
## Screenshots
**Output from running the E2E example showing frequency extraction:**
<img width="1125" height="664" alt="image"
src="https://github.com/user-attachments/assets/455c1ee4-525d-498b-8219-8f12a15292eb"
/>
<img width="1125" height="664" alt="image"
src="https://github.com/user-attachments/assets/64d5da31-85db-427b-b4b4-df47a9c12d6f"
/>
<img width="822" height="456" alt="image"
src="https://github.com/user-attachments/assets/69967354-d550-4818-9aff-a2273e48c5f3"
/>
**Neo4j Browser verification:**
```
✓ Found 6 nodes with frequency_weight in Neo4j!
Sample weighted nodes:
- Weight: 37, Type: ['DocumentChunk']
- Weight: 30, Type: ['Entity']
```
## Pre-submission Checklist
- [x] **I have tested my changes thoroughly before submitting this PR**
- [x] **This PR contains minimal changes necessary to address the
issue/feature**
- [x] My code follows the project's coding standards and style
guidelines
- [x] I have added tests that prove my fix is effective or that my
feature works
- [x] I have added necessary documentation (if applicable)
- [x] All new and existing tests pass
- [x] I have searched existing PRs to ensure this change hasn't been
submitted already
- [x] I have linked any relevant issues in the description
- [x] My commits have clear and descriptive messages
## DCO Affirmation
I affirm that all code in every commit of this pull request conforms to
the terms of the Topoteretes Developer Certificate of Origin.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
## Release Notes
* **New Features**
* Added usage frequency extraction that aggregates interaction data and
weights frequently accessed graph elements.
* Frequency analysis supports configurable time windows, minimum
interaction thresholds, and element type filtering.
* Automatic frequency weight propagation to Neo4j, Kuzu, and generic
graph database backends.
* **Documentation**
* Added comprehensive example script demonstrating end-to-end usage
frequency extraction, weighting, and analysis.
<sub>✏️ Tip: You can customize this high-level summary in your review
settings.</sub>
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
- Remove 3 tests for use_combined_context from test_search.py
- Remove 1 test for use_combined_context from test_search_prepare_search_result_contract.py
- Simplify test_authorized_search_non_combined_delegates to test delegation only
These tests relied on the removed use_combined_context parameter and
CombinedSearchResult type. The functionality is no longer available
after removing combined search support.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Remove use_combined_context parameter from search functions
- Remove CombinedSearchResult class from types module
- Update API routers to remove combined search support
- Remove prepare_combined_context helper function
- Update tutorial notebook to remove use_combined_context usage
- Simplify search return types to always return List[SearchResult]
This removes the combined search feature which aggregated results
across multiple datasets into a single response. Users can still
search across multiple datasets and get results per dataset.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Update all cognify test assertions to expect the chunks_per_batch=None
parameter that was added to the CLI command. This fixes three failing tests:
- test_execute_basic_cognify
- test_cognify_invalid_chunk_size
- test_cognify_nonexistent_ontology_file
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Fix AttributeError when args.chunks_per_batch is not present in the
argparse.Namespace object. Use getattr() with default value of None
to safely access the optional chunks_per_batch parameter.
This resolves test failures in test_cli_edge_cases.py where Namespace
objects were created without the chunks_per_batch attribute.
Changes:
- Use getattr(args, 'chunks_per_batch', None) instead of direct access
- Update test assertion to expect chunks_per_batch=None parameter
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
