Files
nextcloud-mcp-server/.github/workflows/rag-evaluation.yml
T
Chris Coutinho b2742aab80 ci: Add RAG evaluation workflow with workflow_dispatch
Adds a manually-triggered GitHub Actions workflow for RAG evaluation:
- Builds Nextcloud User Manual PDF from documentation source
- Uploads PDF to Nextcloud via WebDAV
- Tags file with 'vector-index' for vector sync indexing
- Waits for vector sync to complete
- Runs RAG integration tests with OpenAI/GitHub Models API

Inputs:
- embedding_model: OpenAI embedding model (default: openai/text-embedding-3-small)
- generation_model: OpenAI generation model (default: openai/gpt-4o-mini)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-11-23 01:22:16 +01:00

272 lines
9.8 KiB
YAML

name: RAG Evaluation
on:
workflow_dispatch:
inputs:
embedding_model:
description: 'OpenAI embedding model'
required: false
default: 'openai/text-embedding-3-small'
generation_model:
description: 'OpenAI generation model'
required: false
default: 'openai/gpt-4o-mini'
jobs:
rag-evaluation:
runs-on: ubuntu-latest
timeout-minutes: 45
steps:
- uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
with:
submodules: 'true'
- name: Clone Nextcloud documentation
uses: actions/checkout@1af3b93b6815bc44a9784bd300feb67ff0d1eeb3 # v6.0.0
with:
repository: 'nextcloud/documentation'
path: 'nextcloud-docs'
- name: Install Sphinx and LaTeX dependencies
run: |
sudo apt-get update
sudo apt-get install -y \
python3-sphinx \
python3-pip \
latexmk \
texlive-latex-recommended \
texlive-latex-extra \
texlive-fonts-recommended \
texlive-fonts-extra
- name: Build User Manual PDF
run: |
cd nextcloud-docs/user_manual
pip3 install -r ../requirements.txt
make latexpdf
ls -la _build/latex/
cp _build/latex/NextcloudUserManual.pdf ../../Nextcloud_User_Manual.pdf
echo "PDF built successfully"
###### Required to build OIDC App ######
- name: Set up php 8.4
uses: shivammathur/setup-php@bf6b4fbd49ca58e4608c9c89fba0b8d90bd2a39f # v2
with:
php-version: 8.4
coverage: none
- name: Install OIDC app composer dependencies
run: |
cd third_party/oidc
composer install --no-dev
###### Required to build OIDC App ######
- name: Run docker compose with vector sync
uses: hoverkraft-tech/compose-action@3846bcd61da338e9eaaf83e7ed0234a12b099b72 # v2.4.1
with:
compose-file: "./docker-compose.yml"
up-flags: "--build"
env:
# Override MCP container environment for OpenAI + vector sync
VECTOR_SYNC_ENABLED: "true"
VECTOR_SYNC_SCAN_INTERVAL: "30"
OPENAI_API_KEY: ${{ secrets.GITHUB_TOKEN }}
OPENAI_BASE_URL: "https://models.github.ai/inference"
OPENAI_EMBEDDING_MODEL: ${{ inputs.embedding_model }}
OPENAI_GENERATION_MODEL: ${{ inputs.generation_model }}
- name: Install the latest version of uv
uses: astral-sh/setup-uv@1e862dfacbd1d6d858c55d9b792c756523627244 # v7.1.4
- name: Wait for Nextcloud to be ready
run: |
echo "Waiting for Nextcloud..."
max_attempts=60
attempt=0
until curl -o /dev/null -s -w "%{http_code}\n" http://localhost:8080/ocs/v2.php/apps/serverinfo/api/v1/info | grep -q "401"; do
attempt=$((attempt + 1))
if [ $attempt -ge $max_attempts ]; then
echo "Service did not become ready in time."
exit 1
fi
echo "Attempt $attempt/$max_attempts: Service not ready, sleeping for 5 seconds..."
sleep 5
done
echo "Nextcloud is ready."
- name: Wait for MCP server to be ready
run: |
echo "Waiting for MCP server..."
max_attempts=30
attempt=0
until curl -o /dev/null -s -w "%{http_code}\n" http://localhost:8000/health | grep -q "200"; do
attempt=$((attempt + 1))
if [ $attempt -ge $max_attempts ]; then
echo "MCP server did not become ready in time."
exit 1
fi
echo "Attempt $attempt/$max_attempts: MCP not ready, sleeping for 2 seconds..."
sleep 2
done
echo "MCP server is ready."
- name: Upload User Manual PDF to Nextcloud
run: |
echo "Uploading Nextcloud_User_Manual.pdf to Nextcloud..."
HTTP_CODE=$(curl -s -o /dev/null -w "%{http_code}" -u admin:admin \
-X PUT \
-T Nextcloud_User_Manual.pdf \
"http://localhost:8080/remote.php/dav/files/admin/Nextcloud_User_Manual.pdf")
if [ "$HTTP_CODE" = "201" ] || [ "$HTTP_CODE" = "204" ]; then
echo "PDF uploaded successfully (HTTP $HTTP_CODE)"
else
echo "Failed to upload PDF (HTTP $HTTP_CODE)"
exit 1
fi
- name: Create vector-index tag
id: create_tag
run: |
# Create the tag using OCS API
echo "Creating vector-index tag..."
RESPONSE=$(curl -s -u admin:admin \
-X POST \
-H 'Content-Type: application/json' \
-H 'OCS-APIRequest: true' \
-d '{"name":"vector-index","userVisible":true,"userAssignable":true}' \
"http://localhost:8080/ocs/v2.php/apps/systemtags/api/v1/tags")
echo "Create tag response: $RESPONSE"
# Get tag ID from response or lookup
TAG_ID=$(echo "$RESPONSE" | grep -oP '(?<="id":)[0-9]+' | head -1 || echo "")
if [ -z "$TAG_ID" ]; then
echo "Tag may already exist, looking it up..."
TAG_ID=$(curl -s -u admin:admin \
-X PROPFIND \
-H 'Content-Type: application/xml' \
-d '<?xml version="1.0"?><d:propfind xmlns:d="DAV:" xmlns:oc="http://owncloud.org/ns"><d:prop><oc:id/><oc:display-name/></d:prop></d:propfind>' \
http://localhost:8080/remote.php/dav/systemtags/ \
| grep -B2 "vector-index" | grep -oP '(?<=<oc:id>)[0-9]+(?=</oc:id>)' | head -1 || echo "")
fi
if [ -z "$TAG_ID" ]; then
echo "ERROR: Could not create or find vector-index tag"
exit 1
fi
echo "Tag ID: $TAG_ID"
echo "tag_id=$TAG_ID" >> $GITHUB_OUTPUT
- name: Get file ID of uploaded PDF
id: get_file_id
run: |
echo "Getting file ID for Nextcloud_User_Manual.pdf..."
# Get file ID using PROPFIND
FILE_ID=$(curl -s -u admin:admin \
-X PROPFIND \
-H 'Content-Type: application/xml' \
-d '<?xml version="1.0"?><d:propfind xmlns:d="DAV:" xmlns:oc="http://owncloud.org/ns"><d:prop><oc:fileid/></d:prop></d:propfind>' \
"http://localhost:8080/remote.php/dav/files/admin/Nextcloud_User_Manual.pdf" \
| grep -oP '(?<=<oc:fileid>)[0-9]+(?=</oc:fileid>)' || echo "")
if [ -z "$FILE_ID" ]; then
echo "ERROR: Could not find file ID"
exit 1
fi
echo "Found file ID: $FILE_ID"
echo "file_id=$FILE_ID" >> $GITHUB_OUTPUT
- name: Tag file with vector-index
env:
FILE_ID: ${{ steps.get_file_id.outputs.file_id }}
TAG_ID: ${{ steps.create_tag.outputs.tag_id }}
run: |
echo "Tagging file $FILE_ID with tag $TAG_ID..."
HTTP_CODE=$(curl -s -o /dev/null -w "%{http_code}" -u admin:admin \
-X PUT \
-H 'Content-Type: application/json' \
-H 'Content-Length: 0' \
"http://localhost:8080/remote.php/dav/systemtags-relations/files/$FILE_ID/$TAG_ID")
if [ "$HTTP_CODE" = "201" ] || [ "$HTTP_CODE" = "409" ]; then
echo "File tagged successfully (HTTP $HTTP_CODE)"
else
echo "Failed to tag file (HTTP $HTTP_CODE)"
exit 1
fi
- name: Wait for vector sync to complete indexing
env:
NEXTCLOUD_HOST: "http://localhost:8080"
NEXTCLOUD_USERNAME: "admin"
NEXTCLOUD_PASSWORD: "admin"
run: |
echo "Waiting for vector sync to index the manual..."
max_attempts=60
attempt=0
# Wait for initial scan to pick up the file
sleep 10
while [ $attempt -lt $max_attempts ]; do
attempt=$((attempt + 1))
# Check vector sync status via MCP
STATUS=$(curl -s http://localhost:8000/health || echo "{}")
echo "Attempt $attempt/$max_attempts: $STATUS"
# Also check indexed count via semantic search
# If we get results, indexing is done
RESULT=$(curl -s -X POST http://localhost:8000/mcp \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"nc_get_vector_sync_status","arguments":{}}}' \
2>/dev/null || echo "{}")
echo "Vector sync status: $RESULT"
# Check if pending is 0 and indexed > 0
INDEXED=$(echo "$RESULT" | jq -r '.result.structuredContent.indexed // 0' 2>/dev/null || echo "0")
PENDING=$(echo "$RESULT" | jq -r '.result.structuredContent.pending // 1' 2>/dev/null || echo "1")
echo "Indexed: $INDEXED, Pending: $PENDING"
if [ "$INDEXED" -gt "0" ] && [ "$PENDING" -eq "0" ]; then
echo "Indexing complete! $INDEXED documents indexed."
break
fi
sleep 10
done
if [ $attempt -ge $max_attempts ]; then
echo "WARNING: Indexing may not be complete, proceeding anyway..."
fi
- name: Run RAG evaluation tests
env:
NEXTCLOUD_HOST: "http://localhost:8080"
NEXTCLOUD_USERNAME: "admin"
NEXTCLOUD_PASSWORD: "admin"
OPENAI_API_KEY: ${{ secrets.GITHUB_TOKEN }}
OPENAI_BASE_URL: "https://models.github.ai/inference"
OPENAI_EMBEDDING_MODEL: ${{ inputs.embedding_model }}
OPENAI_GENERATION_MODEL: ${{ inputs.generation_model }}
run: |
uv run pytest tests/integration/test_rag_openai.py -v --log-cli-level=INFO
- name: Upload test results
if: always()
uses: actions/upload-artifact@v4
with:
name: rag-evaluation-results
path: |
pytest-results.xml
retention-days: 30