Merge branch 'dev' into COG-3146

This commit is contained in:
Vasilije 2026-01-10 08:49:43 +01:00 committed by GitHub
commit f2166be823
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
156 changed files with 9750 additions and 5808 deletions

View file

@ -3,7 +3,7 @@
language: en
early_access: false
enable_free_tier: true
reviews:
reviews:
profile: chill
instructions: >-
# Code Review Instructions
@ -118,10 +118,10 @@ reviews:
- E117
- D208
line_length: 100
dummy_variable_rgx: '^(_.*|junk|extra)$' # Variables starting with '_' or named 'junk' or 'extras', are considered dummy variables
dummy_variable_rgx: '^(_.*|junk|extra)$' # Variables starting with '_' or named 'junk' or 'extras', are considered dummy variables
markdownlint:
enabled: true
yamllint:
enabled: true
chat:
auto_reply: true
auto_reply: true

View file

@ -2,4 +2,8 @@
# Example:
# CORS_ALLOWED_ORIGINS="https://yourdomain.com,https://another.com"
# For local development, you might use:
# CORS_ALLOWED_ORIGINS="http://localhost:3000"
# CORS_ALLOWED_ORIGINS="http://localhost:3000"
LLM_API_KEY="your-openai-api-key"
LLM_MODEL="openai/gpt-4o-mini"
LLM_PROVIDER="openai"

View file

@ -28,4 +28,4 @@ secret-scan:
- path: 'docker-compose.yml'
comment: 'Development docker compose with test credentials (neo4j/pleaseletmein, postgres cognee/cognee)'
- path: 'deployment/helm/docker-compose-helm.yml'
comment: 'Helm deployment docker compose with test postgres credentials (cognee/cognee)'
comment: 'Helm deployment docker compose with test postgres credentials (cognee/cognee)'

View file

@ -8,7 +8,7 @@ body:
attributes:
value: |
Thanks for taking the time to fill out this bug report! Please provide a clear and detailed description.
- type: textarea
id: description
attributes:
@ -17,7 +17,7 @@ body:
placeholder: Describe the bug in detail...
validations:
required: true
- type: textarea
id: reproduction
attributes:
@ -29,7 +29,7 @@ body:
3. See error...
validations:
required: true
- type: textarea
id: expected
attributes:
@ -38,7 +38,7 @@ body:
placeholder: Describe what you expected...
validations:
required: true
- type: textarea
id: actual
attributes:
@ -47,7 +47,7 @@ body:
placeholder: Describe what actually happened...
validations:
required: true
- type: textarea
id: environment
attributes:
@ -61,7 +61,7 @@ body:
- Database: [e.g. Neo4j]
validations:
required: true
- type: textarea
id: logs
attributes:
@ -71,7 +71,7 @@ body:
render: shell
validations:
required: false
- type: textarea
id: additional
attributes:
@ -80,7 +80,7 @@ body:
placeholder: Any additional information...
validations:
required: false
- type: checkboxes
id: checklist
attributes:

View file

@ -8,7 +8,7 @@ body:
attributes:
value: |
Thanks for helping improve our documentation! Please provide details about the documentation issue or improvement.
- type: dropdown
id: doc-type
attributes:
@ -22,7 +22,7 @@ body:
- New documentation request
validations:
required: true
- type: textarea
id: location
attributes:
@ -31,7 +31,7 @@ body:
placeholder: https://cognee.ai/docs/... or specific file/section
validations:
required: true
- type: textarea
id: issue
attributes:
@ -40,7 +40,7 @@ body:
placeholder: The documentation is unclear about...
validations:
required: true
- type: textarea
id: suggestion
attributes:
@ -49,7 +49,7 @@ body:
placeholder: I suggest changing this to...
validations:
required: false
- type: textarea
id: additional
attributes:
@ -58,7 +58,7 @@ body:
placeholder: Additional context...
validations:
required: false
- type: checkboxes
id: checklist
attributes:
@ -71,4 +71,3 @@ body:
required: true
- label: I have specified the location of the documentation issue
required: true

View file

@ -8,7 +8,7 @@ body:
attributes:
value: |
Thanks for suggesting a new feature! Please provide a clear and detailed description of your idea.
- type: textarea
id: problem
attributes:
@ -17,7 +17,7 @@ body:
placeholder: I'm always frustrated when...
validations:
required: true
- type: textarea
id: solution
attributes:
@ -26,7 +26,7 @@ body:
placeholder: I would like to see...
validations:
required: true
- type: textarea
id: alternatives
attributes:
@ -35,7 +35,7 @@ body:
placeholder: I have also considered...
validations:
required: false
- type: textarea
id: use-case
attributes:
@ -44,7 +44,7 @@ body:
placeholder: This feature would help me...
validations:
required: true
- type: textarea
id: implementation
attributes:
@ -53,7 +53,7 @@ body:
placeholder: This could be implemented by...
validations:
required: false
- type: textarea
id: additional
attributes:
@ -62,7 +62,7 @@ body:
placeholder: Additional context...
validations:
required: false
- type: checkboxes
id: checklist
attributes:
@ -75,4 +75,3 @@ body:
required: true
- label: I have described my specific use case
required: true

View file

@ -34,14 +34,14 @@ runs:
-e NEO4J_apoc_export_file_enabled=true \
-e NEO4J_apoc_import_file_enabled=true \
neo4j:${{ inputs.neo4j-version }}
- name: Wait for Neo4j to be ready
shell: bash
run: |
echo "Waiting for Neo4j to start..."
timeout=60
counter=0
while [ $counter -lt $timeout ]; do
if docker exec neo4j-test cypher-shell -u neo4j -p "${{ inputs.neo4j-password }}" "RETURN 1" > /dev/null 2>&1; then
echo "Neo4j is ready!"
@ -51,13 +51,13 @@ runs:
sleep 2
counter=$((counter + 2))
done
if [ $counter -ge $timeout ]; then
echo "Neo4j failed to start within $timeout seconds"
docker logs neo4j-test
exit 1
fi
- name: Verify GDS is available
shell: bash
run: |

View file

@ -8,5 +8,3 @@ lxobr
pazone
siillee
vasilije1990

View file

@ -10,26 +10,21 @@ DO NOT use AI-generated descriptions. We want to understand your thought process
<!--
* 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 -->
## Screenshots
<!-- ADD SCREENSHOT OF LOCAL TESTS PASSING-->
## Pre-submission Checklist
<!-- Please check all boxes that apply before submitting your PR -->
- [ ] **I have tested my changes thoroughly before submitting this PR**
- [ ] **I have tested my changes thoroughly before submitting this PR** (See `CONTRIBUTING.md`)
- [ ] **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

View file

@ -3,7 +3,7 @@ tag-template: 'v$NEXT_PATCH_VERSION'
categories:
- title: 'Features'
labels: ['feature', 'enhancement']
labels: ['feature', 'enhancement']
- title: 'Bug Fixes'
labels: ['bug', 'fix']
- title: 'Maintenance'

View file

@ -34,43 +34,6 @@ env:
ENV: 'dev'
jobs:
lint:
name: Run Linting
runs-on: ubuntu-22.04
steps:
- name: Check out repository
uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Cognee Setup
uses: ./.github/actions/cognee_setup
with:
python-version: ${{ inputs.python-version }}
- name: Run Linting
uses: astral-sh/ruff-action@v2
format-check:
name: Run Formatting Check
runs-on: ubuntu-22.04
steps:
- name: Check out repository
uses: actions/checkout@v4
with:
fetch-depth: 0
- name: Cognee Setup
uses: ./.github/actions/cognee_setup
with:
python-version: ${{ inputs.python-version }}
- name: Run Formatting Check
uses: astral-sh/ruff-action@v2
with:
args: "format --check"
unit-tests:
name: Run Unit Tests
runs-on: ubuntu-22.04

View file

@ -31,54 +31,54 @@ WORKFLOWS=(
for workflow in "${WORKFLOWS[@]}"; do
if [ -f "$workflow" ]; then
echo "Processing $workflow..."
# Create a backup
cp "$workflow" "${workflow}.bak"
# Check if the file begins with a workflow_call trigger
if grep -q "workflow_call:" "$workflow"; then
echo "$workflow already has workflow_call trigger, skipping..."
continue
fi
# Get the content after the 'on:' section
on_line=$(grep -n "^on:" "$workflow" | cut -d ':' -f1)
if [ -z "$on_line" ]; then
echo "Warning: No 'on:' section found in $workflow, skipping..."
continue
fi
# Create a new file with the modified content
{
# Copy the part before 'on:'
head -n $((on_line-1)) "$workflow"
# Add the new on: section that only includes workflow_call
echo "on:"
echo " workflow_call:"
echo " secrets:"
echo " inherit: true"
# Find where to continue after the original 'on:' section
next_section=$(awk "NR > $on_line && /^[a-z]/ {print NR; exit}" "$workflow")
if [ -z "$next_section" ]; then
next_section=$(wc -l < "$workflow")
next_section=$((next_section+1))
fi
# Copy the rest of the file starting from the next section
tail -n +$next_section "$workflow"
} > "${workflow}.new"
# Replace the original with the new version
mv "${workflow}.new" "$workflow"
echo "Modified $workflow to only run when called from test-suites.yml"
else
echo "Warning: $workflow not found, skipping..."
fi
done
echo "Finished modifying workflows!"
echo "Finished modifying workflows!"

View file

@ -45,4 +45,4 @@ jobs:
cache-to: type=registry,ref=cognee/cognee:buildcache,mode=max
- name: Image digest
run: echo ${{ steps.build.outputs.digest }}
run: echo ${{ steps.build.outputs.digest }}

View file

@ -72,5 +72,3 @@ jobs:
} catch (error) {
core.warning(`Failed to add label: ${error.message}`);
}

View file

@ -66,5 +66,3 @@ jobs:
AWS_ACCESS_KEY_ID: ${{ secrets.AWS_S3_DEV_USER_KEY_ID }}
AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_S3_DEV_USER_SECRET_KEY }}
run: uv run python ./cognee/tests/test_load.py

View file

@ -5,7 +5,7 @@ permissions:
contents: read
jobs:
check-uv-lock:
name: Validate uv lockfile and project metadata
name: Lockfile and Pre-commit Hooks
runs-on: ubuntu-22.04
steps:
- name: Check out repository
@ -17,6 +17,9 @@ jobs:
uses: astral-sh/setup-uv@v4
with:
enable-cache: true
- name: Validate uv lockfile and project metadata
run: uv lock --check || { echo "'uv lock --check' failed."; echo "Run 'uv lock' and push your changes."; exit 1; }
- name: Run pre-commit hooks
uses: pre-commit/action@v3.0.1

View file

@ -42,10 +42,10 @@ jobs:
echo "tag=${TAG}" >> "$GITHUB_OUTPUT"
echo "version=${VERSION}" >> "$GITHUB_OUTPUT"
git tag "${TAG}"
git push origin "${TAG}"
- name: Create GitHub Release
uses: softprops/action-gh-release@v2
@ -54,8 +54,8 @@ jobs:
generate_release_notes: true
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
release-pypi-package:
release-pypi-package:
needs: release-github
name: Release PyPI Package from ${{ inputs.flavour }}
permissions:
@ -67,25 +67,25 @@ jobs:
uses: actions/checkout@v4
with:
ref: ${{ inputs.flavour }}
- name: Install uv
uses: astral-sh/setup-uv@v7
- name: Install Python
run: uv python install
- name: Install dependencies
run: uv sync --locked --all-extras
- name: Build distributions
run: uv build
- name: Publish ${{ inputs.flavour }} release to PyPI
env:
UV_PUBLISH_TOKEN: ${{ secrets.PYPI_TOKEN }}
run: uv publish
release-docker-image:
release-docker-image:
needs: release-github
name: Release Docker Image from ${{ inputs.flavour }}
permissions:
@ -128,7 +128,7 @@ jobs:
context: .
platforms: linux/amd64,linux/arm64
push: true
tags: |
tags: |
cognee/cognee:${{ needs.release-github.outputs.version }}
cognee/cognee:latest
labels: |
@ -163,4 +163,4 @@ jobs:
-H "Authorization: Bearer ${{ secrets.REPO_DISPATCH_PAT_TOKEN }}" \
-H "X-GitHub-Api-Version: 2022-11-28" \
https://api.github.com/repos/topoteretes/cognee-community/dispatches \
-d '{"event_type":"new-main-release","client_payload":{"caller_repo":"'"${GITHUB_REPOSITORY}"'"}}'
-d '{"event_type":"new-main-release","client_payload":{"caller_repo":"'"${GITHUB_REPOSITORY}"'"}}'

View file

@ -15,4 +15,3 @@ jobs:
name: Load Tests
uses: ./.github/workflows/load_tests.yml
secrets: inherit

View file

@ -10,7 +10,7 @@ on:
required: false
type: string
default: '["3.10.x", "3.12.x", "3.13.x"]'
os:
os:
required: false
type: string
default: '["ubuntu-22.04", "macos-15", "windows-latest"]'

View file

@ -173,4 +173,4 @@ jobs:
EMBEDDING_MODEL: "amazon.titan-embed-text-v2:0"
EMBEDDING_DIMENSIONS: "1024"
EMBEDDING_MAX_TOKENS: "8191"
run: uv run python ./examples/python/simple_example.py
run: uv run python ./examples/python/simple_example.py

View file

@ -18,11 +18,11 @@ env:
RUNTIME__LOG_LEVEL: ERROR
ENV: 'dev'
jobs:
jobs:
pre-test:
name: basic checks
uses: ./.github/workflows/pre_test.yml
basic-tests:
name: Basic Tests
uses: ./.github/workflows/basic_tests.yml

2
.gitignore vendored
View file

@ -147,6 +147,8 @@ venv/
ENV/
env.bak/
venv.bak/
mise.toml
deployment/helm/values-local.yml
# Spyder project settings
.spyderproject

View file

@ -6,4 +6,4 @@ pull_request_rules:
actions:
backport:
branches:
- main
- main

View file

@ -7,6 +7,7 @@ repos:
- id: trailing-whitespace
- id: end-of-file-fixer
- id: check-yaml
exclude: ^deployment/helm/templates/
- id: check-added-large-files
- repo: https://github.com/astral-sh/ruff-pre-commit
# Ruff version.

View file

@ -128,5 +128,3 @@ MCP server and Frontend:
## CI Mirrors Local Commands
Our GitHub Actions run the same ruff checks and pytest suites shown above (`.github/workflows/basic_tests.yml` and related workflows). Use the commands in this document locally to minimize CI surprises.

View file

@ -1,16 +1,16 @@
> [!IMPORTANT]
> **Note for contributors:** When branching out, create a new branch from the `dev` branch.
# 🎉 Welcome to **cognee**!
# 🎉 Welcome to **cognee**!
We're excited that you're interested in contributing to our project!
We want to ensure that every user and contributor feels welcome, included and supported to participate in cognee community.
We're excited that you're interested in contributing to our project!
We want to ensure that every user and contributor feels welcome, included and supported to participate in cognee community.
This guide will help you get started and ensure your contributions can be efficiently integrated into the project.
## 🌟 Quick Links
- [Code of Conduct](CODE_OF_CONDUCT.md)
- [Discord Community](https://discord.gg/bcy8xFAtfd)
- [Discord Community](https://discord.gg/bcy8xFAtfd)
- [Issue Tracker](https://github.com/topoteretes/cognee/issues)
- [Cognee Docs](https://docs.cognee.ai)
@ -62,6 +62,11 @@ Looking for a place to start? Try filtering for [good first issues](https://gith
## 2. 🛠️ Development Setup
### Required tools
* [Python](https://www.python.org/downloads/)
* [uv](https://docs.astral.sh/uv/getting-started/installation/)
* pre-commit: `uv run pip install pre-commit && pre-commit install`
### Fork and Clone
1. Fork the [**cognee**](https://github.com/topoteretes/cognee) repository
@ -93,14 +98,31 @@ git checkout -b feature/your-feature-name
4. **Commits**: Write clear commit messages
### Running Tests
Rename `.env.example` into `.env` and provide your OPENAI_API_KEY as LLM_API_KEY
```shell
python cognee/cognee/tests/test_library.py
uv run python cognee/tests/test_library.py
```
### Running Simple Example
Rename `.env.example` into `.env` and provide your OPENAI_API_KEY as LLM_API_KEY
Make sure to run ```shell uv sync ``` in the root cloned folder or set up a virtual environment to run cognee
```shell
python examples/python/simple_example.py
```
or
```shell
uv run python examples/python/simple_example.py
```
## 4. 📤 Submitting Changes
1. Install ruff on your system
2. Run ```ruff format .``` and ``` ruff check ``` and fix the issues
1. Make sure that `pre-commit` and hooks are installed. See `Required tools` section for more information. Try executing `pre-commit run` if you are not sure.
3. Push your changes:
```shell
git add .

View file

@ -32,7 +32,7 @@ COPY README.md pyproject.toml uv.lock entrypoint.sh ./
# Install the project's dependencies using the lockfile and settings
RUN --mount=type=cache,target=/root/.cache/uv \
uv sync --extra debug --extra api --extra postgres --extra neo4j --extra llama-index --extra ollama --extra mistral --extra groq --extra anthropic --frozen --no-install-project --no-dev --no-editable
uv sync --extra debug --extra api --extra postgres --extra neo4j --extra llama-index --extra ollama --extra mistral --extra groq --extra anthropic --extra chromadb --frozen --no-install-project --no-dev --no-editable
# Copy Alembic configuration
COPY alembic.ini /app/alembic.ini
@ -43,7 +43,7 @@ COPY alembic/ /app/alembic
COPY ./cognee /app/cognee
COPY ./distributed /app/distributed
RUN --mount=type=cache,target=/root/.cache/uv \
uv sync --extra debug --extra api --extra postgres --extra neo4j --extra llama-index --extra ollama --extra mistral --extra groq --extra anthropic --frozen --no-dev --no-editable
uv sync --extra debug --extra api --extra postgres --extra neo4j --extra llama-index --extra ollama --extra mistral --extra groq --extra anthropic --extra chromadb --frozen --no-dev --no-editable
FROM python:3.12-slim-bookworm

View file

@ -65,12 +65,12 @@ Use your data to build personalized and dynamic memory for AI Agents. Cognee let
## About Cognee
Cognee is an open-source tool and platform that transforms your raw data into persistent and dynamic AI memory for Agents. It combines vector search with graph databases to make your documents both searchable by meaning and connected by relationships.
Cognee is an open-source tool and platform that transforms your raw data into persistent and dynamic AI memory for Agents. It combines vector search with graph databases to make your documents both searchable by meaning and connected by relationships.
You can use Cognee in two ways:
1. [Self-host Cognee Open Source](https://docs.cognee.ai/getting-started/installation), which stores all data locally by default.
2. [Connect to Cognee Cloud](https://platform.cognee.ai/), and get the same OSS stack on managed infrastructure for easier development and productionization.
2. [Connect to Cognee Cloud](https://platform.cognee.ai/), and get the same OSS stack on managed infrastructure for easier development and productionization.
### Cognee Open Source (self-hosted):
@ -81,8 +81,8 @@ You can use Cognee in two ways:
- Offers high customizability through user-defined tasks, modular pipelines, and built-in search endpoints
### Cognee Cloud (managed):
- Hosted web UI dashboard
- Automatic version updates
- Hosted web UI dashboard
- Automatic version updates
- Resource usage analytics
- GDPR compliant, enterprise-grade security
@ -119,7 +119,7 @@ To integrate other LLM providers, see our [LLM Provider Documentation](https://d
### Step 3: Run the Pipeline
Cognee will take your documents, generate a knowledge graph from them and then query the graph based on combined relationships.
Cognee will take your documents, generate a knowledge graph from them and then query the graph based on combined relationships.
Now, run a minimal pipeline:
@ -157,7 +157,7 @@ As you can see, the output is generated from the document we previously stored i
Cognee turns documents into AI memory.
```
### Use the Cognee CLI
### Use the Cognee CLI
As an alternative, you can get started with these essential commands:

View file

@ -1 +1 @@
Generic single-database configuration with an async dbapi.
Generic single-database configuration with an async dbapi.

View file

@ -0,0 +1,52 @@
"""Enable delete for old tutorial notebooks
Revision ID: 1a58b986e6e1
Revises: 46a6ce2bd2b2
Create Date: 2025-12-17 11:04:44.414259
"""
from typing import Sequence, Union
from alembic import op
import sqlalchemy as sa
# revision identifiers, used by Alembic.
revision: str = "1a58b986e6e1"
down_revision: Union[str, None] = "e1ec1dcb50b6"
branch_labels: Union[str, Sequence[str], None] = None
depends_on: Union[str, Sequence[str], None] = None
def change_tutorial_deletable_flag(deletable: bool) -> None:
bind = op.get_bind()
inspector = sa.inspect(bind)
if "notebooks" not in inspector.get_table_names():
return
columns = {col["name"] for col in inspector.get_columns("notebooks")}
required_columns = {"name", "deletable"}
if not required_columns.issubset(columns):
return
notebooks = sa.table(
"notebooks",
sa.Column("name", sa.String()),
sa.Column("deletable", sa.Boolean()),
)
tutorial_name = "Python Development with Cognee Tutorial 🧠"
bind.execute(
notebooks.update().where(notebooks.c.name == tutorial_name).values(deletable=deletable)
)
def upgrade() -> None:
change_tutorial_deletable_flag(True)
def downgrade() -> None:
change_tutorial_deletable_flag(False)

View file

@ -43,10 +43,10 @@ Saiba mais sobre os [casos de uso](https://docs.cognee.ai/use-cases) e [avaliaç
## Funcionalidades
- Conecte e recupere suas conversas passadas, documentos, imagens e transcrições de áudio
- Reduza alucinações, esforço de desenvolvimento e custos
- Carregue dados em bancos de dados de grafos e vetores usando apenas Pydantic
- Transforme e organize seus dados enquanto os coleta de mais de 30 fontes diferentes
- Conecte e recupere suas conversas passadas, documentos, imagens e transcrições de áudio
- Reduza alucinações, esforço de desenvolvimento e custos
- Carregue dados em bancos de dados de grafos e vetores usando apenas Pydantic
- Transforme e organize seus dados enquanto os coleta de mais de 30 fontes diferentes
## Primeiros Passos
@ -108,7 +108,7 @@ if __name__ == '__main__':
Exemplo do output:
```
O Processamento de Linguagem Natural (NLP) é um campo interdisciplinar e transdisciplinar que envolve ciência da computação e recuperação de informações. Ele se concentra na interação entre computadores e a linguagem humana, permitindo que as máquinas compreendam e processem a linguagem natural.
```
Visualização do grafo:

View file

@ -141,7 +141,7 @@ if __name__ == '__main__':
2. Простая демонстрация GraphRAG
[Видео](https://github.com/user-attachments/assets/d80b0776-4eb9-4b8e-aa22-3691e2d44b8f)
3. Cognee с Ollama
3. Cognee с Ollama
[Видео](https://github.com/user-attachments/assets/8621d3e8-ecb8-4860-afb2-5594f2ee17db)
## Правила поведения

View file

@ -114,7 +114,7 @@ if __name__ == '__main__':
示例输出:
```
自然语言处理NLP是计算机科学和信息检索的跨学科领域。它关注计算机和人类语言之间的交互使机器能够理解和处理自然语言。
```
图形可视化:
<a href="https://rawcdn.githack.com/topoteretes/cognee/refs/heads/main/assets/graph_visualization.html"><img src="https://rawcdn.githack.com/topoteretes/cognee/refs/heads/main/assets/graph_visualization.png" width="100%" alt="图形可视化"></a>

File diff suppressed because it is too large Load diff

View file

@ -9,14 +9,15 @@
"lint": "next lint"
},
"dependencies": {
"@auth0/nextjs-auth0": "^4.13.1",
"@auth0/nextjs-auth0": "^4.14.0",
"classnames": "^2.5.1",
"culori": "^4.0.1",
"d3-force-3d": "^3.0.6",
"next": "16.0.4",
"react": "^19.2.0",
"react-dom": "^19.2.0",
"next": "^16.1.7",
"react": "^19.2.3",
"react-dom": "^19.2.3",
"react-force-graph-2d": "^1.27.1",
"react-markdown": "^10.1.0",
"uuid": "^9.0.1"
},
"devDependencies": {

View file

@ -55,7 +55,7 @@ export default function CogneeAddWidget({ onData, useCloud = false }: CogneeAddW
setTrue: setProcessingFilesInProgress,
setFalse: setProcessingFilesDone,
} = useBoolean(false);
const handleAddFiles = (dataset: Dataset, event: ChangeEvent<HTMLInputElement>) => {
event.stopPropagation();

View file

@ -111,7 +111,7 @@ export default function GraphControls({ data, isAddNodeFormOpen, onGraphShapeCha
const [isAuthShapeChangeEnabled, setIsAuthShapeChangeEnabled] = useState(true);
const shapeChangeTimeout = useRef<number | null>(null);
useEffect(() => {
onGraphShapeChange(DEFAULT_GRAPH_SHAPE);

View file

@ -57,7 +57,7 @@ export default function GraphVisualization({ ref, data, graphControls, className
// Initial size calculation
handleResize();
// ResizeObserver
// ResizeObserver
const resizeObserver = new ResizeObserver(() => {
handleResize();
});
@ -216,7 +216,7 @@ export default function GraphVisualization({ ref, data, graphControls, className
}, [data, graphRef]);
const [graphShape, setGraphShape] = useState<string>();
const zoomToFit: ForceGraphMethods["zoomToFit"] = (
durationMs?: number,
padding?: number,
@ -227,15 +227,15 @@ export default function GraphVisualization({ ref, data, graphControls, className
// eslint-disable-next-line @typescript-eslint/no-explicit-any
return undefined as any;
}
return graphRef.current.zoomToFit?.(durationMs, padding, nodeFilter);
};
useImperativeHandle(ref, () => ({
zoomToFit,
setGraphShape,
}));
return (
<div ref={containerRef} className={classNames("w-full h-full", className)} id="graph-container">

View file

@ -1373,4 +1373,4 @@
"padding": 20
}
}
}
}

View file

@ -15,6 +15,8 @@ import AddDataToCognee from "./AddDataToCognee";
import NotebooksAccordion from "./NotebooksAccordion";
import CogneeInstancesAccordion from "./CogneeInstancesAccordion";
import InstanceDatasetsAccordion from "./InstanceDatasetsAccordion";
import cloudFetch from "@/modules/instances/cloudFetch";
import localFetch from "@/modules/instances/localFetch";
interface DashboardProps {
user?: {
@ -26,6 +28,17 @@ interface DashboardProps {
accessToken: string;
}
const cogneeInstances = {
cloudCognee: {
name: "CloudCognee",
fetch: cloudFetch,
},
localCognee: {
name: "LocalCognee",
fetch: localFetch,
}
};
export default function Dashboard({ accessToken }: DashboardProps) {
fetch.setAccessToken(accessToken);
const { user } = useAuthenticatedUser();
@ -38,7 +51,7 @@ export default function Dashboard({ accessToken }: DashboardProps) {
updateNotebook,
saveNotebook,
removeNotebook,
} = useNotebooks();
} = useNotebooks(cogneeInstances.localCognee);
useEffect(() => {
if (!notebooks.length) {

View file

@ -134,7 +134,7 @@ export default function DatasetsAccordion({
} = useBoolean(false);
const [datasetToRemove, setDatasetToRemove] = useState<Dataset | null>(null);
const handleDatasetRemove = (dataset: Dataset) => {
setDatasetToRemove(dataset);
openRemoveDatasetModal();

View file

@ -3,6 +3,7 @@ import { useCallback, useEffect } from "react";
import { fetch, isCloudEnvironment, useBoolean } from "@/utils";
import { checkCloudConnection } from "@/modules/cloud";
import { setApiKey } from "@/modules/instances/cloudFetch";
import { CaretIcon, CloseIcon, CloudIcon, LocalCogneeIcon } from "@/ui/Icons";
import { CTAButton, GhostButton, IconButton, Input, Modal } from "@/ui/elements";
@ -24,6 +25,7 @@ export default function InstanceDatasetsAccordion({ onDatasetsChange }: Instance
const checkConnectionToCloudCognee = useCallback((apiKey?: string) => {
if (apiKey) {
fetch.setApiKey(apiKey);
setApiKey(apiKey);
}
return checkCloudConnection()
.then(setCloudCogneeConnected)

View file

@ -45,7 +45,7 @@ export default function Plan() {
<div className="bg-white rounded-xl px-5 py-5 mb-2">
Affordable and transparent pricing
</div>
<div className="grid grid-cols-3 gap-x-2.5">
<div className="pt-13 py-4 px-5 mb-2.5 rounded-tl-xl rounded-tr-xl bg-white h-full">
<div>Basic</div>

View file

@ -40,7 +40,7 @@ export default function useChat(dataset: Dataset) {
setTrue: disableSearchRun,
setFalse: enableSearchRun,
} = useBoolean(false);
const refreshChat = useCallback(async () => {
const data = await fetchMessages();
return setMessages(data);

View file

@ -46,7 +46,7 @@ function useDatasets(useCloud = false) {
// checkDatasetStatuses(datasets);
// }, 50000);
// }, [fetchDatasetStatuses]);
// useEffect(() => {
// return () => {
// if (statusTimeout.current !== null) {
@ -95,6 +95,7 @@ function useDatasets(useCloud = false) {
})
.catch((error) => {
console.error('Error fetching datasets:', error);
throw error;
});
}, [useCloud]);

View file

@ -0,0 +1,59 @@
import handleServerErrors from "@/utils/handleServerErrors";
// let numberOfRetries = 0;
const cloudApiUrl = process.env.NEXT_PUBLIC_CLOUD_API_URL || "http://localhost:8001";
let apiKey: string | null = process.env.NEXT_PUBLIC_COGWIT_API_KEY || null;
export function setApiKey(newApiKey: string) {
apiKey = newApiKey;
};
export default async function cloudFetch(url: URL | RequestInfo, options: RequestInit = {}): Promise<Response> {
// function retry(lastError: Response) {
// if (numberOfRetries >= 1) {
// return Promise.reject(lastError);
// }
// numberOfRetries += 1;
// return global.fetch("/auth/token")
// .then(() => {
// return fetch(url, options);
// });
// }
const authHeaders = {
"Authorization": `X-Api-Key ${apiKey}`,
};
return global.fetch(
cloudApiUrl + "/api" + (typeof url === "string" ? url : url.toString()).replace("/v1", ""),
{
...options,
headers: {
...options.headers,
...authHeaders,
} as HeadersInit,
credentials: "include",
},
)
.then((response) => handleServerErrors(response, null, true))
.catch((error) => {
if (error.message === "NEXT_REDIRECT") {
throw error;
}
if (error.detail === undefined) {
return Promise.reject(
new Error("No connection to the server.")
);
}
return Promise.reject(error);
});
// .finally(() => {
// numberOfRetries = 0;
// });
}

View file

@ -0,0 +1,27 @@
import handleServerErrors from "@/utils/handleServerErrors";
const localApiUrl = process.env.NEXT_PUBLIC_LOCAL_API_URL || "http://localhost:8000";
export default async function localFetch(url: URL | RequestInfo, options: RequestInit = {}): Promise<Response> {
return global.fetch(
localApiUrl + "/api" + (typeof url === "string" ? url : url.toString()),
{
...options,
credentials: "include",
},
)
.then((response) => handleServerErrors(response, null, false))
.catch((error) => {
if (error.message === "NEXT_REDIRECT") {
throw error;
}
if (error.detail === undefined) {
return Promise.reject(
new Error("No connection to the server.")
);
}
return Promise.reject(error);
});
}

View file

@ -0,0 +1,4 @@
export interface CogneeInstance {
name: string;
fetch: typeof global.fetch;
}

View file

@ -0,0 +1,13 @@
import { CogneeInstance } from "@/modules/instances/types";
export default function createNotebook(notebookName: string, instance: CogneeInstance) {
return instance.fetch("/v1/notebooks/", {
body: JSON.stringify({ name: notebookName }),
method: "POST",
headers: {
"Content-Type": "application/json",
},
}).then((response: Response) =>
response.ok ? response.json() : Promise.reject(response)
);
}

View file

@ -0,0 +1,7 @@
import { CogneeInstance } from "@/modules/instances/types";
export default function deleteNotebook(notebookId: string, instance: CogneeInstance) {
return instance.fetch(`/v1/notebooks/${notebookId}`, {
method: "DELETE",
});
}

View file

@ -0,0 +1,12 @@
import { CogneeInstance } from "@/modules/instances/types";
export default function getNotebooks(instance: CogneeInstance) {
return instance.fetch("/v1/notebooks/", {
method: "GET",
headers: {
"Content-Type": "application/json",
},
}).then((response: Response) =>
response.ok ? response.json() : Promise.reject(response)
);
}

View file

@ -0,0 +1,14 @@
import { Cell } from "@/ui/elements/Notebook/types";
import { CogneeInstance } from "@/modules/instances/types";
export default function runNotebookCell(notebookId: string, cell: Cell, instance: CogneeInstance) {
return instance.fetch(`/v1/notebooks/${notebookId}/${cell.id}/run`, {
body: JSON.stringify({
content: cell.content,
}),
method: "POST",
headers: {
"Content-Type": "application/json",
},
}).then((response: Response) => response.json());
}

View file

@ -0,0 +1,13 @@
import { CogneeInstance } from "@/modules/instances/types";
export default function saveNotebook(notebookId: string, notebookData: object, instance: CogneeInstance) {
return instance.fetch(`/v1/notebooks/${notebookId}`, {
body: JSON.stringify(notebookData),
method: "PUT",
headers: {
"Content-Type": "application/json",
},
}).then((response: Response) =>
response.ok ? response.json() : Promise.reject(response)
);
}

View file

@ -1,20 +1,18 @@
import { useCallback, useState } from "react";
import { fetch, isCloudEnvironment } from "@/utils";
import { Cell, Notebook } from "@/ui/elements/Notebook/types";
import { CogneeInstance } from "@/modules/instances/types";
import createNotebook from "./createNotebook";
import deleteNotebook from "./deleteNotebook";
import getNotebooks from "./getNotebooks";
import runNotebookCell from "./runNotebookCell";
import { default as persistNotebook } from "./saveNotebook";
function useNotebooks() {
function useNotebooks(instance: CogneeInstance) {
const [notebooks, setNotebooks] = useState<Notebook[]>([]);
const addNotebook = useCallback((notebookName: string) => {
return fetch("/v1/notebooks", {
body: JSON.stringify({ name: notebookName }),
method: "POST",
headers: {
"Content-Type": "application/json",
},
}, isCloudEnvironment())
.then((response) => response.json())
.then((notebook) => {
return createNotebook(notebookName, instance)
.then((notebook: Notebook) => {
setNotebooks((notebooks) => [
...notebooks,
notebook,
@ -22,36 +20,29 @@ function useNotebooks() {
return notebook;
});
}, []);
}, [instance]);
const removeNotebook = useCallback((notebookId: string) => {
return fetch(`/v1/notebooks/${notebookId}`, {
method: "DELETE",
}, isCloudEnvironment())
return deleteNotebook(notebookId, instance)
.then(() => {
setNotebooks((notebooks) =>
notebooks.filter((notebook) => notebook.id !== notebookId)
);
});
}, []);
}, [instance]);
const fetchNotebooks = useCallback(() => {
return fetch("/v1/notebooks", {
headers: {
"Content-Type": "application/json",
},
}, isCloudEnvironment())
.then((response) => response.json())
return getNotebooks(instance)
.then((notebooks) => {
setNotebooks(notebooks);
return notebooks;
})
.catch((error) => {
console.error("Error fetching notebooks:", error);
console.error("Error fetching notebooks:", error.detail);
throw error
});
}, []);
}, [instance]);
const updateNotebook = useCallback((updatedNotebook: Notebook) => {
setNotebooks((existingNotebooks) =>
@ -64,20 +55,13 @@ function useNotebooks() {
}, []);
const saveNotebook = useCallback((notebook: Notebook) => {
return fetch(`/v1/notebooks/${notebook.id}`, {
body: JSON.stringify({
name: notebook.name,
cells: notebook.cells,
}),
method: "PUT",
headers: {
"Content-Type": "application/json",
},
}, isCloudEnvironment())
.then((response) => response.json())
}, []);
return persistNotebook(notebook.id, {
name: notebook.name,
cells: notebook.cells,
}, instance);
}, [instance]);
const runCell = useCallback((notebook: Notebook, cell: Cell, cogneeInstance: string) => {
const runCell = useCallback((notebook: Notebook, cell: Cell) => {
setNotebooks((existingNotebooks) =>
existingNotebooks.map((existingNotebook) =>
existingNotebook.id === notebook.id ? {
@ -89,20 +73,11 @@ function useNotebooks() {
error: undefined,
} : existingCell
),
} : notebook
} : existingNotebook
)
);
return fetch(`/v1/notebooks/${notebook.id}/${cell.id}/run`, {
body: JSON.stringify({
content: cell.content,
}),
method: "POST",
headers: {
"Content-Type": "application/json",
},
}, cogneeInstance === "cloud")
.then((response) => response.json())
return runNotebookCell(notebook.id, cell, instance)
.then((response) => {
setNotebooks((existingNotebooks) =>
existingNotebooks.map((existingNotebook) =>
@ -115,11 +90,11 @@ function useNotebooks() {
error: response.error,
} : existingCell
),
} : notebook
} : existingNotebook
)
);
});
}, []);
}, [instance]);
return {
notebooks,

View file

@ -7,4 +7,4 @@ export default function GitHubIcon({ width = 24, height = 24, color = 'currentCo
</g>
</svg>
);
}
}

View file

@ -46,7 +46,7 @@ export default function Header({ user }: HeaderProps) {
checkMCPConnection();
const interval = setInterval(checkMCPConnection, 30000);
return () => clearInterval(interval);
}, [setMCPConnected, setMCPDisconnected]);

View file

@ -90,7 +90,7 @@ export default function SearchView() {
scrollToBottom();
setSearchInputValue("");
// Pass topK to sendMessage
sendMessage(chatInput, searchType, topK)
.then(scrollToBottom)
@ -171,4 +171,4 @@ export default function SearchView() {
</form>
</div>
);
}
}

View file

@ -1,3 +1,2 @@
export { default as Modal } from "./Modal";
export { default as useModal } from "./useModal";

View file

@ -0,0 +1,76 @@
import { memo } from "react";
import ReactMarkdown from "react-markdown";
interface MarkdownPreviewProps {
content: string;
className?: string;
}
function MarkdownPreview({ content, className = "" }: MarkdownPreviewProps) {
return (
<div className={`min-h-24 max-h-96 overflow-y-auto p-4 prose prose-sm max-w-none ${className}`}>
<ReactMarkdown
components={{
h1: ({ children }) => <h1 className="text-2xl font-bold mt-4 mb-2">{children}</h1>,
h2: ({ children }) => <h2 className="text-xl font-bold mt-3 mb-2">{children}</h2>,
h3: ({ children }) => <h3 className="text-lg font-bold mt-3 mb-2">{children}</h3>,
h4: ({ children }) => <h4 className="text-base font-bold mt-2 mb-1">{children}</h4>,
h5: ({ children }) => <h5 className="text-sm font-bold mt-2 mb-1">{children}</h5>,
h6: ({ children }) => <h6 className="text-xs font-bold mt-2 mb-1">{children}</h6>,
p: ({ children }) => <p className="mb-2">{children}</p>,
ul: ({ children }) => <ul className="list-disc list-inside mb-2 ml-4">{children}</ul>,
ol: ({ children }) => <ol className="list-decimal list-inside mb-2 ml-4">{children}</ol>,
li: ({ children }) => <li className="mb-1">{children}</li>,
blockquote: ({ children }) => (
<blockquote className="border-l-4 border-gray-300 pl-4 italic my-2">{children}</blockquote>
),
code: ({ className, children, ...props }) => {
const isInline = !className;
return isInline ? (
<code className="bg-gray-100 px-1 py-0.5 rounded text-sm font-mono" {...props}>
{children}
</code>
) : (
<code className="block bg-gray-100 p-2 rounded text-sm font-mono overflow-x-auto" {...props}>
{children}
</code>
);
},
pre: ({ children }) => (
<pre className="bg-gray-100 p-2 rounded text-sm font-mono overflow-x-auto mb-2">
{children}
</pre>
),
a: ({ href, children }) => (
<a href={href} className="text-blue-600 hover:underline" target="_blank" rel="noopener noreferrer">
{children}
</a>
),
strong: ({ children }) => <strong className="font-bold">{children}</strong>,
em: ({ children }) => <em className="italic">{children}</em>,
hr: () => <hr className="my-4 border-gray-300" />,
table: ({ children }) => (
<div className="overflow-x-auto my-2">
<table className="min-w-full border border-gray-300">{children}</table>
</div>
),
thead: ({ children }) => <thead className="bg-gray-100">{children}</thead>,
tbody: ({ children }) => <tbody>{children}</tbody>,
tr: ({ children }) => <tr className="border-b border-gray-300">{children}</tr>,
th: ({ children }) => (
<th className="border border-gray-300 px-4 py-2 text-left font-bold">
{children}
</th>
),
td: ({ children }) => (
<td className="border border-gray-300 px-4 py-2">{children}</td>
),
}}
>
{content}
</ReactMarkdown>
</div>
);
}
export default memo(MarkdownPreview);

View file

@ -2,15 +2,17 @@
import { v4 as uuid4 } from "uuid";
import classNames from "classnames";
import { Fragment, MouseEvent, RefObject, useCallback, useEffect, useRef, useState } from "react";
import { Fragment, MouseEvent, MutableRefObject, useCallback, useEffect, useRef, useState, memo } from "react";
import { useModal } from "@/ui/elements/Modal";
import { CaretIcon, CloseIcon, PlusIcon } from "@/ui/Icons";
import { IconButton, PopupMenu, TextArea, Modal, GhostButton, CTAButton } from "@/ui/elements";
import PopupMenu from "@/ui/elements/PopupMenu";
import { IconButton, TextArea, Modal, GhostButton, CTAButton } from "@/ui/elements";
import { GraphControlsAPI } from "@/app/(graph)/GraphControls";
import GraphVisualization, { GraphVisualizationAPI } from "@/app/(graph)/GraphVisualization";
import NotebookCellHeader from "./NotebookCellHeader";
import MarkdownPreview from "./MarkdownPreview";
import { Cell, Notebook as NotebookType } from "./types";
interface NotebookProps {
@ -19,7 +21,186 @@ interface NotebookProps {
updateNotebook: (updatedNotebook: NotebookType) => void;
}
interface NotebookCellProps {
cell: Cell;
index: number;
isOpen: boolean;
isMarkdownEditMode: boolean;
onToggleOpen: () => void;
onToggleMarkdownEdit: () => void;
onContentChange: (value: string) => void;
onCellRun: (cell: Cell, cogneeInstance: string) => Promise<void>;
onCellRename: (cell: Cell) => void;
onCellRemove: (cell: Cell) => void;
onCellUp: (cell: Cell) => void;
onCellDown: (cell: Cell) => void;
onCellAdd: (afterCellIndex: number, cellType: "markdown" | "code") => void;
}
const NotebookCell = memo(function NotebookCell({
cell,
index,
isOpen,
isMarkdownEditMode,
onToggleOpen,
onToggleMarkdownEdit,
onContentChange,
onCellRun,
onCellRename,
onCellRemove,
onCellUp,
onCellDown,
onCellAdd,
}: NotebookCellProps) {
return (
<Fragment>
<div className="flex flex-row rounded-xl border-1 border-gray-100">
<div className="flex flex-col flex-1 relative">
{cell.type === "code" ? (
<>
<div className="absolute left-[-1.35rem] top-2.5">
<IconButton className="p-[0.25rem] m-[-0.25rem]" onClick={onToggleOpen}>
<CaretIcon className={classNames("transition-transform", isOpen ? "rotate-0" : "rotate-180")} />
</IconButton>
</div>
<NotebookCellHeader
cell={cell}
runCell={onCellRun}
renameCell={onCellRename}
removeCell={onCellRemove}
moveCellUp={onCellUp}
moveCellDown={onCellDown}
className="rounded-tl-xl rounded-tr-xl"
/>
{isOpen && (
<>
<TextArea
value={cell.content}
onChange={onContentChange}
isAutoExpanding
name="cellInput"
placeholder="Type your code here..."
className="resize-none min-h-36 max-h-96 overflow-y-auto rounded-tl-none rounded-tr-none rounded-bl-xl rounded-br-xl border-0 !outline-0"
/>
<div className="flex flex-col bg-gray-100 overflow-x-auto max-w-full">
{cell.result && (
<div className="px-2 py-2">
output: <CellResult content={cell.result} />
</div>
)}
{!!cell.error?.length && (
<div className="px-2 py-2">
error: {cell.error}
</div>
)}
</div>
</>
)}
</>
) : (
<>
<div className="absolute left-[-1.35rem] top-2.5">
<IconButton className="p-[0.25rem] m-[-0.25rem]" onClick={onToggleOpen}>
<CaretIcon className={classNames("transition-transform", isOpen ? "rotate-0" : "rotate-180")} />
</IconButton>
</div>
<NotebookCellHeader
cell={cell}
renameCell={onCellRename}
removeCell={onCellRemove}
moveCellUp={onCellUp}
moveCellDown={onCellDown}
className="rounded-tl-xl rounded-tr-xl"
/>
{isOpen && (
<div className="relative rounded-tl-none rounded-tr-none rounded-bl-xl rounded-br-xl border-0 overflow-hidden">
<GhostButton
onClick={onToggleMarkdownEdit}
className="absolute top-2 right-2.5 text-xs leading-[1] !px-2 !py-1 !h-auto"
>
{isMarkdownEditMode ? "Preview" : "Edit"}
</GhostButton>
{isMarkdownEditMode ? (
<TextArea
value={cell.content}
onChange={onContentChange}
isAutoExpanding
name="markdownInput"
placeholder="Type your markdown here..."
className="resize-none min-h-24 max-h-96 overflow-y-auto rounded-tl-none rounded-tr-none rounded-bl-xl rounded-br-xl border-0 !outline-0 !bg-gray-50"
/>
) : (
<MarkdownPreview content={cell.content} className="!bg-gray-50" />
)}
</div>
)}
</>
)}
</div>
</div>
<div className="ml-[-1.35rem]">
<PopupMenu
openToRight={true}
triggerElement={<PlusIcon />}
triggerClassName="p-[0.25rem] m-[-0.25rem]"
>
<div className="flex flex-col gap-0.5">
<button
onClick={() => onCellAdd(index, "markdown")}
className="hover:bg-gray-100 w-full text-left px-2 cursor-pointer"
>
<span>text</span>
</button>
</div>
<div
onClick={() => onCellAdd(index, "code")}
className="hover:bg-gray-100 w-full text-left px-2 cursor-pointer"
>
<span>code</span>
</div>
</PopupMenu>
</div>
</Fragment>
);
});
export default function Notebook({ notebook, updateNotebook, runCell }: NotebookProps) {
const [openCells, setOpenCells] = useState(new Set(notebook.cells.map((c: Cell) => c.id)));
const [markdownEditMode, setMarkdownEditMode] = useState<Set<string>>(new Set());
const toggleCellOpen = useCallback((id: string) => {
setOpenCells((prev) => {
const newState = new Set(prev);
if (newState.has(id)) {
newState.delete(id)
} else {
newState.add(id);
}
return newState;
});
}, []);
const toggleMarkdownEditMode = useCallback((id: string) => {
setMarkdownEditMode((prev) => {
const newState = new Set(prev);
if (newState.has(id)) {
newState.delete(id);
} else {
newState.add(id);
}
return newState;
});
}, []);
useEffect(() => {
if (notebook.cells.length === 0) {
const newCell: Cell = {
@ -34,7 +215,7 @@ export default function Notebook({ notebook, updateNotebook, runCell }: Notebook
});
toggleCellOpen(newCell.id)
}
}, [notebook, updateNotebook]);
}, [notebook, updateNotebook, toggleCellOpen]);
const handleCellRun = useCallback((cell: Cell, cogneeInstance: string) => {
return runCell(notebook, cell, cogneeInstance);
@ -43,7 +224,7 @@ export default function Notebook({ notebook, updateNotebook, runCell }: Notebook
const handleCellAdd = useCallback((afterCellIndex: number, cellType: "markdown" | "code") => {
const newCell: Cell = {
id: uuid4(),
name: "new cell",
name: cellType === "markdown" ? "Markdown Cell" : "Code Cell",
type: cellType,
content: "",
};
@ -59,7 +240,7 @@ export default function Notebook({ notebook, updateNotebook, runCell }: Notebook
toggleCellOpen(newCell.id);
updateNotebook(newNotebook);
}, [notebook, updateNotebook]);
}, [notebook, updateNotebook, toggleCellOpen]);
const removeCell = useCallback((cell: Cell, event?: MouseEvent) => {
event?.preventDefault();
@ -81,14 +262,12 @@ export default function Notebook({ notebook, updateNotebook, runCell }: Notebook
openCellRemoveConfirmModal(cell);
}, [openCellRemoveConfirmModal]);
const handleCellInputChange = useCallback((notebook: NotebookType, cell: Cell, value: string) => {
const newCell = {...cell, content: value };
const handleCellInputChange = useCallback((cellId: string, value: string) => {
updateNotebook({
...notebook,
cells: notebook.cells.map((cell: Cell) => (cell.id === newCell.id ? newCell : cell)),
cells: notebook.cells.map((cell: Cell) => (cell.id === cellId ? {...cell, content: value} : cell)),
});
}, [updateNotebook]);
}, [notebook, updateNotebook]);
const handleCellUp = useCallback((cell: Cell) => {
const index = notebook.cells.indexOf(cell);
@ -131,133 +310,28 @@ export default function Notebook({ notebook, updateNotebook, runCell }: Notebook
}
}, [notebook, updateNotebook]);
const [openCells, setOpenCells] = useState(new Set(notebook.cells.map((c: Cell) => c.id)));
const toggleCellOpen = (id: string) => {
setOpenCells((prev) => {
const newState = new Set(prev);
if (newState.has(id)) {
newState.delete(id)
} else {
newState.add(id);
}
return newState;
});
};
return (
<>
<div className="bg-white rounded-xl flex flex-col gap-0.5 px-7 py-5 flex-1">
<div className="mb-5">{notebook.name}</div>
{notebook.cells.map((cell: Cell, index) => (
<Fragment key={cell.id}>
<div key={cell.id} className="flex flex-row rounded-xl border-1 border-gray-100">
<div className="flex flex-col flex-1 relative">
{cell.type === "code" ? (
<>
<div className="absolute left-[-1.35rem] top-2.5">
<IconButton className="p-[0.25rem] m-[-0.25rem]" onClick={toggleCellOpen.bind(null, cell.id)}>
<CaretIcon className={classNames("transition-transform", openCells.has(cell.id) ? "rotate-0" : "rotate-180")} />
</IconButton>
</div>
<NotebookCellHeader
cell={cell}
runCell={handleCellRun}
renameCell={handleCellRename}
removeCell={handleCellRemove}
moveCellUp={handleCellUp}
moveCellDown={handleCellDown}
className="rounded-tl-xl rounded-tr-xl"
/>
{openCells.has(cell.id) && (
<>
<TextArea
value={cell.content}
onChange={handleCellInputChange.bind(null, notebook, cell)}
// onKeyUp={handleCellRunOnEnter}
isAutoExpanding
name="cellInput"
placeholder="Type your code here..."
contentEditable={true}
className="resize-none min-h-36 max-h-96 overflow-y-auto rounded-tl-none rounded-tr-none rounded-bl-xl rounded-br-xl border-0 !outline-0"
/>
<div className="flex flex-col bg-gray-100 overflow-x-auto max-w-full">
{cell.result && (
<div className="px-2 py-2">
output: <CellResult content={cell.result} />
</div>
)}
{!!cell.error?.length && (
<div className="px-2 py-2">
error: {cell.error}
</div>
)}
</div>
</>
)}
</>
) : (
<>
<div className="absolute left-[-1.35rem] top-2.5">
<IconButton className="p-[0.25rem] m-[-0.25rem]" onClick={toggleCellOpen.bind(null, cell.id)}>
<CaretIcon className={classNames("transition-transform", openCells.has(cell.id) ? "rotate-0" : "rotate-180")} />
</IconButton>
</div>
<NotebookCellHeader
cell={cell}
renameCell={handleCellRename}
removeCell={handleCellRemove}
moveCellUp={handleCellUp}
moveCellDown={handleCellDown}
className="rounded-tl-xl rounded-tr-xl"
/>
{openCells.has(cell.id) && (
<TextArea
value={cell.content}
onChange={handleCellInputChange.bind(null, notebook, cell)}
// onKeyUp={handleCellRunOnEnter}
isAutoExpanding
name="cellInput"
placeholder="Type your text here..."
contentEditable={true}
className="resize-none min-h-24 max-h-96 overflow-y-auto rounded-tl-none rounded-tr-none rounded-bl-xl rounded-br-xl border-0 !outline-0"
/>
)}
</>
)}
</div>
</div>
<div className="ml-[-1.35rem]">
<PopupMenu
openToRight={true}
triggerElement={<PlusIcon />}
triggerClassName="p-[0.25rem] m-[-0.25rem]"
>
<div className="flex flex-col gap-0.5">
<button
onClick={() => handleCellAdd(index, "markdown")}
className="hover:bg-gray-100 w-full text-left px-2 cursor-pointer"
>
<span>text</span>
</button>
</div>
<div
onClick={() => handleCellAdd(index, "code")}
className="hover:bg-gray-100 w-full text-left px-2 cursor-pointer"
>
<span>code</span>
</div>
</PopupMenu>
</div>
</Fragment>
<NotebookCell
key={cell.id}
cell={cell}
index={index}
isOpen={openCells.has(cell.id)}
isMarkdownEditMode={markdownEditMode.has(cell.id)}
onToggleOpen={() => toggleCellOpen(cell.id)}
onToggleMarkdownEdit={() => toggleMarkdownEditMode(cell.id)}
onContentChange={(value) => handleCellInputChange(cell.id, value)}
onCellRun={handleCellRun}
onCellRename={handleCellRename}
onCellRemove={handleCellRemove}
onCellUp={handleCellUp}
onCellDown={handleCellDown}
onCellAdd={handleCellAdd}
/>
))}
</div>
@ -288,6 +362,10 @@ function CellResult({ content }: { content: [] }) {
getSelectedNode: () => null,
});
if (content.length === 0) {
return <span>OK</span>;
}
for (const line of content) {
try {
if (Array.isArray(line)) {
@ -298,7 +376,7 @@ function CellResult({ content }: { content: [] }) {
<span className="text-sm pl-2 mb-4">reasoning graph</span>
<GraphVisualization
data={transformInsightsGraphData(line)}
ref={graphRef as RefObject<GraphVisualizationAPI>}
ref={graphRef as MutableRefObject<GraphVisualizationAPI>}
graphControls={graphControls}
className="min-h-80"
/>
@ -346,7 +424,7 @@ function CellResult({ content }: { content: [] }) {
<span className="text-sm pl-2 mb-4">reasoning graph (datasets: {datasetName})</span>
<GraphVisualization
data={transformToVisualizationData(graph)}
ref={graphRef as RefObject<GraphVisualizationAPI>}
ref={graphRef as MutableRefObject<GraphVisualizationAPI>}
graphControls={graphControls}
className="min-h-80"
/>
@ -356,8 +434,7 @@ function CellResult({ content }: { content: [] }) {
}
}
}
if (typeof(line) === "object" && line["result"] && typeof(line["result"]) === "string") {
else if (typeof(line) === "object" && line["result"] && typeof(line["result"]) === "string") {
const datasets = Array.from(
// eslint-disable-next-line @typescript-eslint/no-explicit-any
new Set(Object.values(line["datasets"]).map((dataset: any) => dataset.name))
@ -369,39 +446,46 @@ function CellResult({ content }: { content: [] }) {
<span className="block px-2 py-2 whitespace-normal">{line["result"]}</span>
</div>
);
}
if (typeof(line) === "object" && line["graphs"]) {
Object.entries<{ nodes: []; edges: []; }>(line["graphs"]).forEach(([datasetName, graph]) => {
parsedContent.push(
<div key={datasetName} className="w-full h-full bg-white">
<span className="text-sm pl-2 mb-4">reasoning graph (datasets: {datasetName})</span>
<GraphVisualization
data={transformToVisualizationData(graph)}
ref={graphRef as RefObject<GraphVisualizationAPI>}
graphControls={graphControls}
className="min-h-80"
/>
</div>
);
});
}
if (typeof(line) === "object" && line["result"] && typeof(line["result"]) === "object") {
if (line["graphs"]) {
Object.entries<{ nodes: []; edges: []; }>(line["graphs"]).forEach(([datasetName, graph]) => {
parsedContent.push(
<div key={datasetName} className="w-full h-full bg-white">
<span className="text-sm pl-2 mb-4">reasoning graph (datasets: {datasetName})</span>
<GraphVisualization
data={transformToVisualizationData(graph)}
ref={graphRef as MutableRefObject<GraphVisualizationAPI>}
graphControls={graphControls}
className="min-h-80"
/>
</div>
);
});
}
}
else if (typeof(line) === "object" && line["result"] && typeof(line["result"]) === "object") {
parsedContent.push(
<pre className="px-2 w-full h-full bg-white text-sm" key={String(line).slice(0, -10)}>
{JSON.stringify(line["result"], null, 2)}
</pre>
)
}
if (typeof(line) === "string") {
else if (typeof(line) === "object") {
parsedContent.push(
<pre className="px-2 w-full h-full bg-white text-sm" key={String(line).slice(0, -10)}>
{JSON.stringify(line, null, 2)}
</pre>
)
}
else if (typeof(line) === "string") {
parsedContent.push(
<pre className="px-2 w-full h-full bg-white text-sm whitespace-normal" key={String(line).slice(0, -10)}>
{line}
</pre>
)
}
} catch (error) {
console.error(error);
} catch {
// It is fine if we don't manage to parse the output line, we show it as it is.
parsedContent.push(
<pre className="px-2 w-full h-full bg-white text-sm whitespace-normal" key={String(line).slice(0, -10)}>
{line}
@ -415,7 +499,6 @@ function CellResult({ content }: { content: [] }) {
{item}
</div>
));
};
function transformToVisualizationData(graph: { nodes: [], edges: [] }) {
@ -451,7 +534,7 @@ function transformInsightsGraphData(triplets: Triplet[]) {
target: string,
label: string,
}
} = {};
} = {};
for (const triplet of triplets) {
nodes[triplet[0].id] = {
@ -471,7 +554,7 @@ function transformInsightsGraphData(triplets: Triplet[]) {
label: triplet[1]["relationship_name"],
};
}
return {
nodes: Object.values(nodes),
links: Object.values(links),

View file

@ -1,9 +1,12 @@
"use client";
import { useState } from "react";
import classNames from "classnames";
import { isCloudEnvironment, useBoolean } from "@/utils";
import { PlayIcon } from "@/ui/Icons";
import { PopupMenu, IconButton } from "@/ui/elements";
import PopupMenu from "@/ui/elements/PopupMenu";
import { IconButton } from "@/ui/elements";
import { LoadingIndicator } from "@/ui/App";
import { Cell } from "./types";
@ -39,7 +42,7 @@ export default function NotebookCellHeader({
if (runCell) {
setIsRunningCell();
runCell(cell, runInstance)
.then(() => {
.finally(() => {
setIsNotRunningCell();
});
}
@ -53,7 +56,7 @@ export default function NotebookCellHeader({
{isRunningCell ? <LoadingIndicator /> : <IconButton onClick={handleCellRun}><PlayIcon /></IconButton>}
</>
)}
<span className="ml-4">{cell.name}</span>
<span className="ml-4">{cell.type === "markdown" ? "Markdown Cell" : cell.name}</span>
</div>
<div className="pr-4 flex flex-row items-center gap-8">
{runCell && (

View file

@ -1,12 +1,12 @@
"use client";
import classNames from "classnames";
import { InputHTMLAttributes, useCallback, useEffect, useLayoutEffect, useRef } from "react"
import { InputHTMLAttributes, useCallback, useEffect, useRef } from "react"
interface TextAreaProps extends Omit<InputHTMLAttributes<HTMLTextAreaElement>, "onChange"> {
isAutoExpanding?: boolean; // Set to true to enable auto-expanding text area behavior. Default is false.
value: string;
onChange: (value: string) => void;
value?: string;
onChange?: (value: string) => void;
}
export default function TextArea({
@ -19,95 +19,81 @@ export default function TextArea({
placeholder = "",
onKeyUp,
...props
}: TextAreaProps) {
const handleTextChange = useCallback((event: Event) => {
const fakeTextAreaElement = event.target as HTMLDivElement;
const newValue = fakeTextAreaElement.innerText;
}: TextAreaProps) {
const textareaRef = useRef<HTMLTextAreaElement>(null);
const maxHeightRef = useRef<number | null>(null);
const throttleTimeoutRef = useRef<number | null>(null);
const lastAdjustTimeRef = useRef<number>(0);
const THROTTLE_MS = 250; // 4 calculations per second
const adjustHeight = useCallback(() => {
if (!isAutoExpanding || !textareaRef.current) return;
const textarea = textareaRef.current;
// Cache maxHeight on first calculation
if (maxHeightRef.current === null) {
const computedStyle = getComputedStyle(textarea);
maxHeightRef.current = computedStyle.maxHeight === "none"
? Infinity
: parseInt(computedStyle.maxHeight) || Infinity;
}
// Reset height to auto to get the correct scrollHeight
textarea.style.height = "auto";
// Set height to scrollHeight, but respect max-height
const scrollHeight = textarea.scrollHeight;
textarea.style.height = `${Math.min(scrollHeight, maxHeightRef.current)}px`;
lastAdjustTimeRef.current = Date.now();
}, [isAutoExpanding]);
const handleChange = useCallback((event: React.ChangeEvent<HTMLTextAreaElement>) => {
const newValue = event.target.value;
onChange?.(newValue);
}, [onChange]);
const handleKeyUp = useCallback((event: Event) => {
if (onKeyUp) {
onKeyUp(event as unknown as React.KeyboardEvent<HTMLTextAreaElement>);
}
}, [onKeyUp]);
// Throttle height adjustments to avoid blocking typing
if (isAutoExpanding) {
const now = Date.now();
const timeSinceLastAdjust = now - lastAdjustTimeRef.current;
const handleTextAreaFocus = (event: React.FocusEvent<HTMLDivElement>) => {
if (event.target.innerText.trim() === placeholder) {
event.target.innerText = "";
}
};
const handleTextAreaBlur = (event: React.FocusEvent<HTMLDivElement>) => {
if (value === "") {
event.target.innerText = placeholder;
}
};
const handleChange = (event: React.ChangeEvent<HTMLTextAreaElement>) => {
onChange(event.target.value);
};
const fakeTextAreaRef = useRef<HTMLDivElement>(null);
useLayoutEffect(() => {
const fakeTextAreaElement = fakeTextAreaRef.current;
if (fakeTextAreaElement && fakeTextAreaElement.innerText.trim() !== "") {
fakeTextAreaElement.innerText = placeholder;
}
}, [placeholder]);
useLayoutEffect(() => {
const fakeTextAreaElement = fakeTextAreaRef.current;
if (fakeTextAreaElement) {
fakeTextAreaElement.addEventListener("input", handleTextChange);
fakeTextAreaElement.addEventListener("keyup", handleKeyUp);
}
return () => {
if (fakeTextAreaElement) {
fakeTextAreaElement.removeEventListener("input", handleTextChange);
fakeTextAreaElement.removeEventListener("keyup", handleKeyUp);
if (timeSinceLastAdjust >= THROTTLE_MS) {
adjustHeight();
} else {
if (throttleTimeoutRef.current !== null) {
clearTimeout(throttleTimeoutRef.current);
}
throttleTimeoutRef.current = window.setTimeout(() => {
adjustHeight();
throttleTimeoutRef.current = null;
}, THROTTLE_MS - timeSinceLastAdjust);
}
};
}, [handleKeyUp, handleTextChange]);
}
}, [onChange, isAutoExpanding, adjustHeight]);
useEffect(() => {
const fakeTextAreaElement = fakeTextAreaRef.current;
const textAreaText = fakeTextAreaElement?.innerText;
if (fakeTextAreaElement && (value === "" || value === "\n")) {
fakeTextAreaElement.innerText = placeholder;
return;
if (isAutoExpanding && textareaRef.current) {
adjustHeight();
}
}, [value, isAutoExpanding, adjustHeight]);
if (fakeTextAreaElement && textAreaText !== value) {
fakeTextAreaElement.innerText = value;
}
}, [placeholder, value]);
useEffect(() => {
return () => {
if (throttleTimeoutRef.current !== null) {
clearTimeout(throttleTimeoutRef.current);
}
};
}, []);
return isAutoExpanding ? (
<>
<div
ref={fakeTextAreaRef}
contentEditable="true"
role="textbox"
aria-multiline="true"
className={classNames("block w-full rounded-md bg-white px-4 py-4 text-base text-gray-900 outline-1 -outline-offset-1 outline-gray-300 placeholder:text-gray-400 focus:outline-2 focus:-outline-offset-2 focus:outline-indigo-600", className)}
onFocus={handleTextAreaFocus}
onBlur={handleTextAreaBlur}
/>
</>
) : (
return (
<textarea
ref={isAutoExpanding ? textareaRef : undefined}
name={name}
style={style}
value={value}
placeholder={placeholder}
className={classNames("block w-full rounded-md bg-white px-4 py-4 text-base text-gray-900 outline-1 -outline-offset-1 outline-gray-300 placeholder:text-gray-400 focus:outline-2 focus:-outline-offset-2 focus:outline-indigo-600", className)}
onChange={handleChange}
onKeyUp={onKeyUp}
{...props}
/>
)

View file

@ -10,4 +10,4 @@ export { default as NeutralButton } from "./NeutralButton";
export { default as StatusIndicator } from "./StatusIndicator";
export { default as StatusDot } from "./StatusDot";
export { default as Accordion } from "./Accordion";
export { default as Notebook } from "./Notebook";
export { default as Notebook } from "./Notebook";

View file

@ -57,7 +57,7 @@ export default async function fetch(url: string, options: RequestInit = {}, useC
new Error("Backend server is not responding. Please check if the server is running.")
);
}
if (error.detail === undefined) {
return Promise.reject(
new Error("No connection to the server.")
@ -74,7 +74,7 @@ export default async function fetch(url: string, options: RequestInit = {}, useC
fetch.checkHealth = async () => {
const maxRetries = 5;
const retryDelay = 1000; // 1 second
for (let i = 0; i < maxRetries; i++) {
try {
const response = await global.fetch(`${backendApiUrl.replace("/api", "")}/health`);
@ -90,7 +90,7 @@ fetch.checkHealth = async () => {
await new Promise(resolve => setTimeout(resolve, retryDelay));
}
}
throw new Error("Backend server is not responding after multiple attempts");
};

View file

@ -1,8 +1,12 @@
import { redirect } from "next/navigation";
export default function handleServerErrors(response: Response, retry?: (response: Response) => Promise<Response>, useCloud?: boolean): Promise<Response> {
export default function handleServerErrors(
response: Response,
retry: ((response: Response) => Promise<Response>) | null = null,
useCloud: boolean = false,
): Promise<Response> {
return new Promise((resolve, reject) => {
if (response.status === 401 && !useCloud) {
if ((response.status === 401 || response.status === 403) && !useCloud) {
if (retry) {
return retry(response)
.catch(() => {

View file

@ -105,14 +105,14 @@ If you'd rather run cognee-mcp in a container, you have two options:
```bash
# For HTTP transport (recommended for web deployments)
docker run -e TRANSPORT_MODE=http --env-file ./.env -p 8000:8000 --rm -it cognee/cognee-mcp:main
# For SSE transport
# For SSE transport
docker run -e TRANSPORT_MODE=sse --env-file ./.env -p 8000:8000 --rm -it cognee/cognee-mcp:main
# For stdio transport (default)
docker run -e TRANSPORT_MODE=stdio --env-file ./.env --rm -it cognee/cognee-mcp:main
```
**Installing optional dependencies at runtime:**
You can install optional dependencies when running the container by setting the `EXTRAS` environment variable:
```bash
# Install a single optional dependency group at runtime
@ -122,7 +122,7 @@ If you'd rather run cognee-mcp in a container, you have two options:
--env-file ./.env \
-p 8000:8000 \
--rm -it cognee/cognee-mcp:main
# Install multiple optional dependency groups at runtime (comma-separated)
docker run \
-e TRANSPORT_MODE=sse \
@ -131,7 +131,7 @@ If you'd rather run cognee-mcp in a container, you have two options:
-p 8000:8000 \
--rm -it cognee/cognee-mcp:main
```
**Available optional dependency groups:**
- `aws` - S3 storage support
- `postgres` / `postgres-binary` - PostgreSQL database support
@ -160,7 +160,7 @@ If you'd rather run cognee-mcp in a container, you have two options:
# With stdio transport (default)
docker run -e TRANSPORT_MODE=stdio --env-file ./.env --rm -it cognee/cognee-mcp:main
```
**With runtime installation of optional dependencies:**
```bash
# Install optional dependencies from Docker Hub image
@ -357,7 +357,7 @@ You can configure both transports simultaneously for testing:
"url": "http://localhost:8000/sse"
},
"cognee-http": {
"type": "http",
"type": "http",
"url": "http://localhost:8000/mcp"
}
}

View file

@ -7,11 +7,11 @@ echo "Environment: $ENVIRONMENT"
# Install optional dependencies if EXTRAS is set
if [ -n "$EXTRAS" ]; then
echo "Installing optional dependencies: $EXTRAS"
# Get the cognee version that's currently installed
COGNEE_VERSION=$(uv pip show cognee | grep "Version:" | awk '{print $2}')
echo "Current cognee version: $COGNEE_VERSION"
# Build the extras list for cognee
IFS=',' read -ra EXTRA_ARRAY <<< "$EXTRAS"
# Combine base extras from pyproject.toml with requested extras
@ -28,11 +28,11 @@ if [ -n "$EXTRAS" ]; then
fi
fi
done
echo "Installing cognee with extras: $ALL_EXTRAS"
echo "Running: uv pip install 'cognee[$ALL_EXTRAS]==$COGNEE_VERSION'"
uv pip install "cognee[$ALL_EXTRAS]==$COGNEE_VERSION"
# Verify installation
echo ""
echo "✓ Optional dependencies installation completed"
@ -93,19 +93,19 @@ if [ -n "$API_URL" ]; then
if echo "$API_URL" | grep -q "localhost" || echo "$API_URL" | grep -q "127.0.0.1"; then
echo "⚠️ Warning: API_URL contains localhost/127.0.0.1"
echo " Original: $API_URL"
# Try to use host.docker.internal (works on Mac/Windows and recent Linux with Docker Desktop)
FIXED_API_URL=$(echo "$API_URL" | sed 's/localhost/host.docker.internal/g' | sed 's/127\.0\.0\.1/host.docker.internal/g')
echo " Converted to: $FIXED_API_URL"
echo " This will work on Mac/Windows/Docker Desktop."
echo " On Linux without Docker Desktop, you may need to:"
echo " - Use --network host, OR"
echo " - Set API_URL=http://172.17.0.1:8000 (Docker bridge IP)"
API_URL="$FIXED_API_URL"
fi
API_ARGS="--api-url $API_URL"
if [ -n "$API_TOKEN" ]; then
API_ARGS="$API_ARGS --api-token $API_TOKEN"

View file

@ -16,4 +16,4 @@ EMBEDDING_API_VERSION=""
GRAPHISTRY_USERNAME=""
GRAPHISTRY_PASSWORD=""
GRAPHISTRY_PASSWORD=""

View file

@ -14,7 +14,7 @@ This starter kit is deprecated. Its examples have been integrated into the `/new
# Cognee Starter Kit
Welcome to the <a href="https://github.com/topoteretes/cognee">cognee</a> Starter Repo! This repository is designed to help you get started quickly by providing a structured dataset and pre-built data pipelines using cognee to build powerful knowledge graphs.
You can use this repo to ingest, process, and visualize data in minutes.
You can use this repo to ingest, process, and visualize data in minutes.
By following this guide, you will:
@ -80,7 +80,7 @@ Custom model uses custom pydantic model for graph extraction. This script catego
python src/pipelines/custom-model.py
```
## Graph preview
## Graph preview
cognee provides a visualize_graph function that will render the graph for you.

View file

@ -71,7 +71,7 @@ def get_sync_router() -> APIRouter:
-H "Content-Type: application/json" \\
-H "Cookie: auth_token=your-token" \\
-d '{"dataset_ids": ["123e4567-e89b-12d3-a456-426614174000", "456e7890-e12b-34c5-d678-901234567000"]}'
# Sync all user datasets (empty request body or null dataset_ids)
curl -X POST "http://localhost:8000/api/v1/sync" \\
-H "Content-Type: application/json" \\
@ -88,7 +88,7 @@ def get_sync_router() -> APIRouter:
- **413 Payload Too Large**: Dataset too large for current cloud plan
- **429 Too Many Requests**: Rate limit exceeded
## Notes
## Notes
- Sync operations run in the background - you get an immediate response
- Use the returned run_id to track progress (status API coming soon)
- Large datasets are automatically chunked for efficient transfer
@ -179,7 +179,7 @@ def get_sync_router() -> APIRouter:
```
## Example Responses
**No running syncs:**
```json
{

View file

@ -21,7 +21,7 @@ binary streams, then stores them in a specified dataset for further processing.
Supported Input Types:
- **Text strings**: Direct text content
- **File paths**: Local file paths (absolute paths starting with "/")
- **File paths**: Local file paths (absolute paths starting with "/")
- **File URLs**: "file:///absolute/path" or "file://relative/path"
- **S3 paths**: "s3://bucket-name/path/to/file"
- **Lists**: Multiple files or text strings in a single call

View file

@ -17,7 +17,7 @@ The `cognee config` command allows you to view and modify configuration settings
You can:
- View all current configuration settings
- Get specific configuration values
- Get specific configuration values
- Set configuration values
- Unset (reset to default) specific configuration values
- Reset all configuration to defaults

View file

@ -290,7 +290,7 @@ class NeptuneAnalyticsAdapter(NeptuneGraphDB, VectorDBInterface):
query_string = f"""
CALL neptune.algo.vectors.topKByEmbeddingWithFiltering({{
topK: {limit},
embedding: {embedding},
embedding: {embedding},
nodeFilter: {{ equals: {{property: '{self._COLLECTION_PREFIX}', value: '{collection_name}'}} }}
}}
)
@ -299,7 +299,7 @@ class NeptuneAnalyticsAdapter(NeptuneGraphDB, VectorDBInterface):
if with_vector:
query_string += """
WITH node, score, id(node) as node_id
WITH node, score, id(node) as node_id
MATCH (n)
WHERE id(n) = id(node)
CALL neptune.algo.vectors.get(n)

View file

@ -14,6 +14,8 @@ from tenacity import (
)
import litellm
import os
from urllib.parse import urlparse
import httpx
from cognee.infrastructure.databases.vector.embeddings.EmbeddingEngine import EmbeddingEngine
from cognee.infrastructure.databases.exceptions import EmbeddingException
from cognee.infrastructure.llm.tokenizer.HuggingFace import (
@ -79,10 +81,26 @@ class LiteLLMEmbeddingEngine(EmbeddingEngine):
enable_mocking = str(enable_mocking).lower()
self.mock = enable_mocking in ("true", "1", "yes")
# Validate provided custom embedding endpoint early to avoid long hangs later
if self.endpoint:
try:
parsed = urlparse(self.endpoint)
except Exception:
parsed = None
if not parsed or parsed.scheme not in ("http", "https") or not parsed.netloc:
logger.error(
"Invalid EMBEDDING_ENDPOINT configured: '%s'. Expected a URL starting with http:// or https://",
str(self.endpoint),
)
raise EmbeddingException(
"Invalid EMBEDDING_ENDPOINT. Please set a valid URL (e.g., https://host:port) "
"via environment variable EMBEDDING_ENDPOINT."
)
@retry(
stop=stop_after_delay(128),
stop=stop_after_delay(30),
wait=wait_exponential_jitter(2, 128),
retry=retry_if_not_exception_type(litellm.exceptions.NotFoundError),
retry=retry_if_not_exception_type((litellm.exceptions.NotFoundError, EmbeddingException)),
before_sleep=before_sleep_log(logger, logging.DEBUG),
reraise=True,
)
@ -111,12 +129,16 @@ class LiteLLMEmbeddingEngine(EmbeddingEngine):
return [data["embedding"] for data in response["data"]]
else:
async with embedding_rate_limiter_context_manager():
response = await litellm.aembedding(
model=self.model,
input=text,
api_key=self.api_key,
api_base=self.endpoint,
api_version=self.api_version,
# Ensure each attempt does not hang indefinitely
response = await asyncio.wait_for(
litellm.aembedding(
model=self.model,
input=text,
api_key=self.api_key,
api_base=self.endpoint,
api_version=self.api_version,
),
timeout=30.0,
)
return [data["embedding"] for data in response.data]
@ -154,6 +176,27 @@ class LiteLLMEmbeddingEngine(EmbeddingEngine):
logger.error("Context window exceeded for embedding text: %s", str(error))
raise error
except asyncio.TimeoutError as e:
# Per-attempt timeout likely an unreachable endpoint
logger.error(
"Embedding endpoint timed out. EMBEDDING_ENDPOINT='%s'. "
"Verify that the endpoint is reachable and correct.",
str(self.endpoint),
)
raise EmbeddingException(
"Embedding request timed out. Check EMBEDDING_ENDPOINT connectivity."
) from e
except (httpx.ConnectError, httpx.ReadTimeout) as e:
logger.error(
"Failed to connect to embedding endpoint. EMBEDDING_ENDPOINT='%s'. "
"Ensure the URL is correct and the server is running.",
str(self.endpoint),
)
raise EmbeddingException(
"Cannot connect to embedding endpoint. Check EMBEDDING_ENDPOINT."
) from e
except (
litellm.exceptions.BadRequestError,
litellm.exceptions.NotFoundError,
@ -162,8 +205,15 @@ class LiteLLMEmbeddingEngine(EmbeddingEngine):
raise EmbeddingException(f"Failed to index data points using model {self.model}") from e
except Exception as error:
logger.error("Error embedding text: %s", str(error))
raise error
# Fall back to a clear, actionable message for connectivity/misconfiguration issues
logger.error(
"Error embedding text: %s. EMBEDDING_ENDPOINT='%s'.",
str(error),
str(self.endpoint),
)
raise EmbeddingException(
"Embedding failed due to an unexpected error. Verify EMBEDDING_ENDPOINT and provider settings."
) from error
def get_vector_size(self) -> int:
"""

View file

@ -10,4 +10,4 @@ Extraction rules:
5. Current-time references ("now", "current", "today"): If the query explicitly refers to the present, set both starts_at and ends_at to now (the ingestion timestamp).
6. "Who is" and "Who was" questions: These imply a general identity or biographical inquiry without a specific temporal scope. Set both starts_at and ends_at to None.
7. Ordering rule: Always ensure the earlier date is assigned to starts_at and the later date to ends_at.
8. No temporal information: If no valid or inferable time reference is found, set both starts_at and ends_at to None.
8. No temporal information: If no valid or inferable time reference is found, set both starts_at and ends_at to None.

View file

@ -22,4 +22,4 @@ The `attributes` should be a list of dictionaries, each containing:
- Relationships should be technical with one or at most two words. If two words, use underscore camelcase style
- Relationships could imply general meaning like: subject, object, participant, recipient, agent, instrument, tool, source, cause, effect, purpose, manner, resource, etc.
- You can combine two words to form a relationship name: subject_role, previous_owner, etc.
- Focus on how the entity specifically relates to the event
- Focus on how the entity specifically relates to the event

View file

@ -27,4 +27,4 @@ class Event(BaseModel):
time_from: Optional[Timestamp] = None
time_to: Optional[Timestamp] = None
location: Optional[str] = None
```
```

View file

@ -19,8 +19,8 @@ The aim is to achieve simplicity and clarity in the knowledge graph.
- **Naming Convention**: Use snake_case for relationship names, e.g., `acted_in`.
# 3. Coreference Resolution
- **Maintain Entity Consistency**: When extracting entities, it's vital to ensure consistency.
If an entity, such as "John Doe", is mentioned multiple times in the text but is referred to by different names or pronouns (e.g., "Joe", "he"),
always use the most complete identifier for that entity throughout the knowledge graph. In this example, use "John Doe" as the Persons ID.
If an entity, is mentioned multiple times in the text but is referred to by different names or pronouns,
always use the most complete identifier for that entity throughout the knowledge graph.
Remember, the knowledge graph should be coherent and easily understandable, so maintaining consistency in entity references is crucial.
# 4. Strict Compliance
Adhere to the rules strictly. Non-compliance will result in termination

View file

@ -22,7 +22,7 @@ You are an advanced algorithm designed to extract structured information to buil
3. **Coreference Resolution**:
- Maintain one consistent node ID for each real-world entity.
- Resolve aliases, acronyms, and pronouns to the most complete form.
- *Example*: Always use "John Doe" even if later referred to as "Doe" or "he".
- *Example*: Always use full identifier even if later referred to as in a similar but slightly different way
**Property & Data Guidelines**:

View file

@ -42,10 +42,10 @@ You are an advanced algorithm designed to extract structured information from un
- **Rule**: Resolve all aliases, acronyms, and pronouns to one canonical identifier.
> **One-Shot Example**:
> **Input**: "John Doe is an author. Later, Doe published a book. He is well-known."
> **Input**: "X is an author. Later, Doe published a book. He is well-known."
> **Output Node**:
> ```
> John Doe (Person)
> X (Person)
> ```
---

View file

@ -15,7 +15,7 @@ You are an advanced algorithm that extracts structured data into a knowledge gra
- Properties are key-value pairs; do not use escaped quotes.
3. **Coreference Resolution**
- Use a single, complete identifier for each entity (e.g., always "John Doe" not "Joe" or "he").
- Use a single, complete identifier for each entity
4. **Relationship Labels**:
- Use descriptive, lowercase, snake_case names for edges.

View file

@ -26,7 +26,7 @@ Use **basic atomic types** for node labels. Always prefer general types over spe
- Good: "Alan Turing", "Google Inc.", "World War II"
- Bad: "Entity_001", "1234", "he", "they"
- Never use numeric or autogenerated IDs.
- Prioritize **most complete form** of entity names for consistency (e.g., always use "John Doe" instead of "John" or "he").
- Prioritize **most complete form** of entity names for consistency
2. Dates, Numbers, and Properties
---------------------------------

View file

@ -2,12 +2,12 @@ You are an expert query analyzer for a **GraphRAG system**. Your primary goal is
Here are the available `SearchType` tools and their specific functions:
- **`SUMMARIES`**: The `SUMMARIES` search type retrieves summarized information from the knowledge graph.
- **`SUMMARIES`**: The `SUMMARIES` search type retrieves summarized information from the knowledge graph.
**Best for:**
**Best for:**
- Getting concise overviews of topics
- Summarizing large amounts of information
- Getting concise overviews of topics
- Summarizing large amounts of information
- Quick understanding of complex subjects
**Best for:**
@ -16,7 +16,7 @@ Here are the available `SearchType` tools and their specific functions:
- Understanding relationships between concepts
- Exploring the structure of your knowledge graph
* **`CHUNKS`**: The `CHUNKS` search type retrieves specific facts and information chunks from the knowledge graph.
* **`CHUNKS`**: The `CHUNKS` search type retrieves specific facts and information chunks from the knowledge graph.
**Best for:**
@ -122,4 +122,4 @@ Response: `NATURAL_LANGUAGE`
Your response MUST be a single word, consisting of only the chosen `SearchType` name. Do not provide any explanation.
Your response MUST be a single word, consisting of only the chosen `SearchType` name. Do not provide any explanation.

View file

@ -1 +1 @@
Respond with: test
Respond with: test

View file

@ -34,6 +34,7 @@ class LLMProvider(Enum):
GEMINI = "gemini"
MISTRAL = "mistral"
BEDROCK = "bedrock"
LLAMA_CPP = "llama_cpp"
def get_llm_client(raise_api_key_error: bool = True):
@ -187,5 +188,28 @@ def get_llm_client(raise_api_key_error: bool = True):
instructor_mode=llm_config.llm_instructor_mode.lower(),
)
elif provider == LLMProvider.LLAMA_CPP:
from cognee.infrastructure.llm.structured_output_framework.litellm_instructor.llm.llama_cpp.adapter import (
LlamaCppAPIAdapter,
)
# Get optional local mode parameters (will be None if not set)
# TODO: refactor llm_config to include these parameters, currently they cannot be defined and defaults are used
model_path = getattr(llm_config, "llama_cpp_model_path", None)
n_ctx = getattr(llm_config, "llama_cpp_n_ctx", 2048)
n_gpu_layers = getattr(llm_config, "llama_cpp_n_gpu_layers", 0)
chat_format = getattr(llm_config, "llama_cpp_chat_format", "chatml")
return LlamaCppAPIAdapter(
model=llm_config.llm_model,
max_completion_tokens=max_completion_tokens,
instructor_mode=llm_config.llm_instructor_mode.lower(),
endpoint=llm_config.llm_endpoint,
api_key=llm_config.llm_api_key,
model_path=model_path,
n_ctx=n_ctx,
n_gpu_layers=n_gpu_layers,
chat_format=chat_format,
)
else:
raise UnsupportedLLMProviderError(provider)

View file

@ -0,0 +1,191 @@
"""Adapter for Instructor-backed Structured Output Framework for Llama CPP"""
import litellm
import logging
import instructor
from typing import Type, Optional
from openai import AsyncOpenAI
from pydantic import BaseModel
from cognee.infrastructure.llm.structured_output_framework.litellm_instructor.llm.llm_interface import (
LLMInterface,
)
from cognee.shared.logging_utils import get_logger
from cognee.shared.rate_limiting import llm_rate_limiter_context_manager
from tenacity import (
retry,
stop_after_delay,
wait_exponential_jitter,
retry_if_not_exception_type,
before_sleep_log,
)
logger = get_logger()
class LlamaCppAPIAdapter(LLMInterface):
"""
Adapter for Llama CPP LLM provider with support for TWO modes:
1. SERVER MODE (OpenAI-compatible):
- Connects to llama-cpp-python server via HTTP (local or remote)
- Uses instructor.from_openai()
- Requires: endpoint, api_key, model
2. LOCAL MODE (In-process):
- Loads model directly using llama-cpp-python library
- Uses instructor.patch() on llama.Llama object
- Requires: model_path
Public methods:
- acreate_structured_output
Instance variables:
- name
- model (for server mode) or model_path (for local mode)
- mode_type: "server" or "local"
- max_completion_tokens
- aclient
"""
name: str
model: Optional[str]
model_path: Optional[str]
mode_type: str # "server" or "local"
default_instructor_mode = instructor.Mode.JSON
def __init__(
self,
name: str = "LlamaCpp",
max_completion_tokens: int = 2048,
instructor_mode: Optional[str] = None,
# Server mode parameters
endpoint: Optional[str] = None,
api_key: Optional[str] = None,
model: Optional[str] = None,
# Local mode parameters
model_path: Optional[str] = None,
n_ctx: int = 2048,
n_gpu_layers: int = 0,
chat_format: str = "chatml",
):
self.name = name
self.max_completion_tokens = max_completion_tokens
self.instructor_mode = instructor_mode if instructor_mode else self.default_instructor_mode
# Determine which mode to use
if model_path:
self._init_local_mode(model_path, n_ctx, n_gpu_layers, chat_format)
elif endpoint:
self._init_server_mode(endpoint, api_key, model)
else:
raise ValueError(
"Must provide either 'model_path' (for local mode) or 'endpoint' (for server mode)"
)
def _init_local_mode(self, model_path: str, n_ctx: int, n_gpu_layers: int, chat_format: str):
"""Initialize local mode using llama-cpp-python library directly"""
try:
import llama_cpp
except ImportError:
raise ImportError(
"llama-cpp-python is not installed. Install with: pip install llama-cpp-python"
)
logger.info(f"Initializing LlamaCpp in LOCAL mode with model: {model_path}")
self.mode_type = "local"
self.model_path = model_path
self.model = None
# Initialize llama-cpp-python with the model
self.llama = llama_cpp.Llama(
model_path=model_path,
n_gpu_layers=n_gpu_layers, # -1 for all GPU, 0 for CPU only
chat_format=chat_format,
n_ctx=n_ctx,
verbose=False,
)
self.aclient = instructor.patch(
create=self.llama.create_chat_completion_openai_v1,
mode=instructor.Mode(self.instructor_mode),
)
def _init_server_mode(self, endpoint: str, api_key: Optional[str], model: Optional[str]):
"""Initialize server mode connecting to llama-cpp-python server"""
logger.info(f"Initializing LlamaCpp in SERVER mode with endpoint: {endpoint}")
self.mode_type = "server"
self.model = model
self.model_path = None
self.endpoint = endpoint
self.api_key = api_key
# Use instructor.from_openai() for server mode (OpenAI-compatible API)
self.aclient = instructor.from_openai(
AsyncOpenAI(base_url=self.endpoint, api_key=self.api_key),
mode=instructor.Mode(self.instructor_mode),
)
@retry(
stop=stop_after_delay(128),
wait=wait_exponential_jitter(8, 128),
retry=retry_if_not_exception_type(litellm.exceptions.NotFoundError),
before_sleep=before_sleep_log(logger, logging.DEBUG),
reraise=True,
)
async def acreate_structured_output(
self, text_input: str, system_prompt: str, response_model: Type[BaseModel], **kwargs
) -> BaseModel:
"""
Generate a structured output from the LLM using the provided text and system prompt.
Works in both local and server modes transparently.
Parameters:
-----------
- text_input (str): The input text provided by the user.
- system_prompt (str): The system prompt that guides the response generation.
- response_model (Type[BaseModel]): The model type that the response should conform to.
Returns:
--------
- BaseModel: A structured output that conforms to the specified response model.
"""
async with llm_rate_limiter_context_manager():
# Prepare messages (system first, then user is more standard)
messages = [
{"role": "system", "content": system_prompt},
{"role": "user", "content": text_input},
]
if self.mode_type == "server":
# Server mode: use async client with OpenAI-compatible API
response = await self.aclient.chat.completions.create(
model=self.model,
messages=messages,
response_model=response_model,
max_retries=2,
max_completion_tokens=self.max_completion_tokens,
**kwargs,
)
else:
import asyncio
# Local mode: instructor.patch() returns a SYNC callable
# Per docs: https://python.useinstructor.com/integrations/llama-cpp-python/
def _call_sync():
return self.aclient(
messages=messages,
response_model=response_model,
max_tokens=self.max_completion_tokens,
**kwargs,
)
# Run sync function in thread pool to avoid blocking
response = await asyncio.to_thread(_call_sync)
return response

View file

@ -3,3 +3,4 @@ from .get_notebooks import get_notebooks
from .create_notebook import create_notebook
from .update_notebook import update_notebook
from .delete_notebook import delete_notebook
from .create_tutorial_notebooks import create_tutorial_notebooks

View file

@ -6,40 +6,6 @@ from cognee.infrastructure.databases.relational import with_async_session
from ..models.Notebook import Notebook, NotebookCell
TUTORIAL_NOTEBOOK_NAME = "Python Development with Cognee Tutorial 🧠"
async def _create_tutorial_notebook(
user_id: UUID, session: AsyncSession, force_refresh: bool = False
) -> None:
"""
Create the default tutorial notebook for new users.
Dynamically fetches from: https://github.com/topoteretes/cognee/blob/notebook_tutorial/notebooks/starter_tutorial.zip
"""
TUTORIAL_ZIP_URL = (
"https://github.com/topoteretes/cognee/raw/notebook_tutorial/notebooks/starter_tutorial.zip"
)
try:
# Create notebook from remote zip file (includes notebook + data files)
notebook = await Notebook.from_ipynb_zip_url(
zip_url=TUTORIAL_ZIP_URL,
owner_id=user_id,
notebook_filename="tutorial.ipynb",
name=TUTORIAL_NOTEBOOK_NAME,
deletable=False,
force=force_refresh,
)
# Add to session and commit
session.add(notebook)
await session.commit()
except Exception as e:
print(f"Failed to fetch tutorial notebook from {TUTORIAL_ZIP_URL}: {e}")
raise e
@with_async_session
async def create_notebook(

View file

@ -0,0 +1,191 @@
from pathlib import Path
from uuid import NAMESPACE_OID, UUID, uuid5, uuid4
from typing import List, Optional, Dict, Any
import re
import json
from sqlalchemy.ext.asyncio import AsyncSession
from cognee.shared.logging_utils import get_logger
from cognee.root_dir import ROOT_DIR
from ..models.Notebook import Notebook, NotebookCell
logger = get_logger()
def _get_tutorials_directory() -> Path:
"""Get the path to the tutorials directory."""
return ROOT_DIR / "modules" / "notebooks" / "tutorials"
def _parse_cell_index(filename: str) -> int:
"""Extract cell index from filename like 'cell-0.md' or 'cell-123.py'."""
match = re.search(r"cell-(\d+)", filename)
if match:
return int(match.group(1))
return -1
def _get_cell_type(file_path: Path) -> str:
"""Determine cell type from file extension."""
extension = file_path.suffix.lower()
if extension == ".md":
return "markdown"
elif extension == ".py":
return "code"
else:
raise ValueError(f"Unsupported cell file type: {extension}")
def _extract_markdown_heading(content: str) -> str | None:
"""Extract the first markdown heading from content."""
for line in content.splitlines():
line = line.strip()
# Match lines starting with one or more # followed by space and text
match = re.match(r"^#+\s+(.+)$", line)
if match:
return match.group(1).strip()
return None
def _get_cell_name(cell_file: Path, cell_type: str, content: str) -> str:
"""Get the appropriate name for a cell."""
if cell_type == "code":
return "Code Cell"
elif cell_type == "markdown":
heading = _extract_markdown_heading(content)
if heading:
return heading
# Fallback to filename stem
return cell_file.stem
def _load_tutorial_cells(tutorial_dir: Path) -> List[NotebookCell]:
"""Load all cells from a tutorial directory, sorted by cell index."""
cells = []
cell_files = [
file_path
for file_path in tutorial_dir.iterdir()
if file_path.is_file()
and file_path.name.startswith("cell-")
and file_path.suffix in [".md", ".py"]
]
cell_files.sort(key=lambda file_path: _parse_cell_index(file_path.name))
for cell_file in cell_files:
try:
cell_type = _get_cell_type(cell_file)
content = cell_file.read_text(encoding="utf-8")
cell_name = _get_cell_name(cell_file, cell_type, content)
cells.append(
NotebookCell(
id=uuid4(),
type=cell_type,
name=cell_name,
content=content,
)
)
except Exception as e:
logger.warning(f"Failed to load cell {cell_file}: {e}")
continue
return cells
def _read_tutorial_config(tutorial_dir: Path) -> Optional[Dict[str, Any]]:
"""Read config.json from a tutorial directory if it exists."""
config_path = tutorial_dir / "config.json"
if config_path.exists():
try:
with open(config_path, "r", encoding="utf-8") as f:
return json.load(f)
except (json.JSONDecodeError, IOError) as e:
logger.warning(f"Failed to read config.json from {tutorial_dir}: {e}")
return None
return None
def _format_tutorial_name(tutorial_dir_name: str) -> str:
"""Format tutorial directory name into a readable notebook name (fallback)."""
name = tutorial_dir_name.replace("-", " ").replace("_", " ")
return f"{name.capitalize()} - tutorial 🧠"
async def create_tutorial_notebooks(user_id: UUID, session: AsyncSession) -> None:
"""
Create tutorial notebooks for all tutorials found in the tutorials directory.
Each tutorial directory will become a separate notebook.
"""
try:
tutorials_dir = _get_tutorials_directory()
if not tutorials_dir.exists():
logger.warning(f"Tutorials directory not found: {tutorials_dir}")
return
tutorial_dirs = [
d for d in tutorials_dir.iterdir() if d.is_dir() and not d.name.startswith(".")
]
if not tutorial_dirs:
logger.warning(f"No tutorial directories found in {tutorials_dir}")
return
notebooks_to_add = []
for tutorial_dir in tutorial_dirs:
try:
cells = _load_tutorial_cells(tutorial_dir)
if not cells:
logger.warning(f"No cells found in tutorial directory: {tutorial_dir}")
continue
config = _read_tutorial_config(tutorial_dir)
# Use name from config.json, or fallback to formatted directory name
if config and "name" in config:
notebook_name = config["name"]
else:
notebook_name = _format_tutorial_name(tutorial_dir.name)
logger.warning(
f"No config.json or 'name' field found in {tutorial_dir}, "
f"using fallback name: {notebook_name}"
)
# Use deletable flag from config.json, or default to False for tutorials
deletable = False
if config and "deletable" in config:
deletable = bool(config["deletable"])
notebook_id = uuid5(NAMESPACE_OID, name=notebook_name)
notebook = Notebook(
id=notebook_id,
owner_id=user_id,
name=notebook_name,
cells=cells,
deletable=deletable,
)
notebooks_to_add.append(notebook)
logger.info(f"Created tutorial notebook: {notebook_name} with {len(cells)} cells")
except Exception as e:
logger.error(f"Failed to create tutorial notebook from {tutorial_dir}: {e}")
continue
if not notebooks_to_add:
return
for notebook in notebooks_to_add:
session.add(notebook)
await session.commit()
except Exception as e:
logger.error(f"Failed to create tutorial notebooks for user {user_id}: {e}")

View file

@ -1,4 +1,4 @@
from uuid import UUID
from uuid import NAMESPACE_OID, UUID, uuid5
from typing import List
from sqlalchemy import select, and_
from sqlalchemy.ext.asyncio import AsyncSession
@ -6,7 +6,7 @@ from sqlalchemy.ext.asyncio import AsyncSession
from cognee.infrastructure.databases.relational import with_async_session
from ..models.Notebook import Notebook
from .create_notebook import _create_tutorial_notebook, TUTORIAL_NOTEBOOK_NAME
from .create_tutorial_notebooks import create_tutorial_notebooks
from cognee.shared.logging_utils import get_logger
@ -19,21 +19,25 @@ async def get_notebooks(
session: AsyncSession,
) -> List[Notebook]:
# Check if tutorial notebook already exists for this user
tutorial_notebook_ids = [
uuid5(NAMESPACE_OID, name="Cognee Basics - tutorial 🧠"),
uuid5(NAMESPACE_OID, name="Python Development with Cognee - tutorial 🧠"),
]
tutorial_query = select(Notebook).where(
and_(
Notebook.owner_id == user_id,
Notebook.name == TUTORIAL_NOTEBOOK_NAME,
Notebook.id.in_(tutorial_notebook_ids),
~Notebook.deletable,
)
)
tutorial_result = await session.execute(tutorial_query)
tutorial_notebook = tutorial_result.scalar_one_or_none()
tutorial_notebooks = tutorial_result.scalars().all()
# If tutorial notebook doesn't exist, create it
if tutorial_notebook is None:
logger.info(f"Tutorial notebook not found for user {user_id}, creating it")
# If tutorial notebooks don't exist, create them
if len(tutorial_notebooks) == 0:
logger.info(f"Tutorial notebooks not found for user {user_id}, creating them")
try:
await _create_tutorial_notebook(user_id, session, force_refresh=False)
await create_tutorial_notebooks(user_id, session)
except Exception as e:
# Log the error but continue to return existing notebooks
logger.error(f"Failed to create tutorial notebook for user {user_id}: {e}")

View file

@ -0,0 +1,3 @@
# Use Cognee to Build Your Own Knowledge Graph
Cognee is a tool that allows you to build your own knowledge graph from the data you have.

View file

@ -0,0 +1,10 @@
# What You'll Learn in This Tutorial
In this tutorial, you'll learn how to use Cognee to transform scattered data into an intelligent knowledge system that enhances your workflow.
By the end, you'll have:
- Connected disparate data sources into a unified AI memory graph
- Built a memory layer that infers knowledge from provided data
- Learn how to use search capabilities that combine the diverse context
This tutorial demonstrates the power of knowledge graphs and retrieval-augmented generation (RAG), showing you how to build systems that learn from data and infer knowledge.

View file

@ -0,0 +1,7 @@
# Cognee and Its Core Operations
Before we dive in, let's understand the core Cognee operations we'll be working with:
- `cognee.add()` - Ingests raw data into the system
- `cognee.cognify()` - Processes and structures data into a knowledge graph using AI
- `cognee.search()` - Queries the knowledge graph with natural language

View file

@ -0,0 +1,28 @@
# Add data one by one, or pass a list to add multiple items at once
await cognee.add(
"Harry Potter is a student at Hogwarts and belongs to Gryffindor house. \
He is known for defeating Voldemort and his Patronus is a stag.",
dataset_name="cognee-basics",
)
await cognee.add(
"Hermione Granger is a student at Hogwarts and also belongs to Gryffindor house. \
She is known for her intelligence and deep knowledge of spells. Her Patronus is an otter.",
dataset_name="cognee-basics",
)
await cognee.add(
"Severus Snape is a professor at Hogwarts who teaches Potions. \
He belongs to Slytherin house and was secretly loyal to Albus Dumbledore.",
dataset_name="cognee-basics",
)
await cognee.add(
[
"Hogwarts is a magical school located in Scotland. During Harry Potter's time at school, the headmaster was Albus Dumbledore.",
"A Horcrux is a dark magic object used to store a fragment of a wizard's soul. Voldemort created multiple Horcruxes to achieve immortality.",
"The Elder Wand is a powerful wand believed to be unbeatable. Its final known owner was Harry Potter.",
],
dataset_name="cognee-basics",
)

View file

@ -0,0 +1,3 @@
# Run cognify to process the data and create a knowledge graph
await cognee.cognify(datasets=["cognee-basics"])

View file

@ -0,0 +1,9 @@
# And after the cognification, search the knowledge graph
result = await cognee.search(
"Which characters belong to Gryffindor?",
datasets=["cognee-basics"],
)
# Print the result so you can see it in the notebook output
print(result)

View file

@ -0,0 +1,17 @@
# Run multiple searches and print the results
result_1 = await cognee.search(
"Who taught Potions at Hogwarts at time Albus Dumbledore was the headmaster?",
datasets=["cognee-basics"],
)
# Print the result so you can see it in the notebook output
print(result_1)
result_2 = await cognee.search(
"How to defeat Voldemort?",
datasets=["cognee-basics"],
)
print(result_2)

View file

@ -0,0 +1,4 @@
{
"name": "Cognee Basics - tutorial 🧠",
"deletable": false
}

View file

@ -0,0 +1,3 @@
# Using Cognee with Python Development Data
Unite authoritative Python practice (Guido van Rossum's own contributions!), normative guidance (Zen/PEP8), and your lived context (rules + conversations) into one AI memory that produces answers that are relevant, explainable, and consistent.

View file

@ -0,0 +1,3 @@
You'll see that cognee has connected your Python development challenges with Guido's approaches, revealing patterns like:
- "Type hint implementation failed due to circular imports - similar to issue Guido solved in mypy PR #1234"
- "Performance bottleneck in list comprehension matches pattern Guido optimized in CPython commit abc123"

Some files were not shown because too many files have changed in this diff Show more