Integrate with GitHub Actions

The MemVault GitHub Action allows your CI/CD pipelines to automatically index repository documentation, store deployment history, and retrieve relevant context from past builds.

Why Use MemVault in CI/CD?

Modern software development generates massive amounts of context:

Deployment logs and outcomes
Test results and failure patterns
Code review discussions
Documentation changes
Performance metrics

Instead of losing this valuable information after each run, MemVault stores it in a searchable, semantically-aware memory system. Your pipelines can then retrieve relevant historical context to make smarter decisions.

Prerequisites

Before you start, you need:

A MemVault account with an active API key
A GitHub repository with Actions enabled
Basic familiarity with GitHub Actions workflow syntax

Installation

Step 1: Get Your API Key

Sign up at memvault-demo.vercel.app
Complete checkout and check your email for your API key
Your key will start with sk_

Step 2: Add API Key to GitHub Secrets

Go to your repository on GitHub
Click Settings → Secrets and variables → Actions
Click New repository secret
Name: MEMVAULT_API_KEY
Value: Paste your API key
Click Add secret

Step 3: Create Workflow File

Create .github/workflows/memvault-sync.yml in your repository:

name: Sync to MemVault
on:
  push:
    branches: [ main ]
jobs:
  sync:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: MemVault Sync
        uses: jakops88-hub/long-term-memory-api/.github/actions/memvault-sync@v1
        with:
          memvault_api_key: ${{ secrets.MEMVAULT_API_KEY }}

That's it. Every time you push to main, your repository documentation will be synced to MemVault.

Configuration Options

The memvault-sync action supports several configuration options:

- name: MemVault Sync
  uses: jakops88-hub/long-term-memory-api/.github/actions/memvault-sync@v1
  with:
    # Required: Your MemVault API key
    memvault_api_key: ${{ secrets.MEMVAULT_API_KEY }}
    
    # Optional: Custom user ID (defaults to repository name)
    user_id: my-org-repos
    
    # Optional: Custom agent ID (defaults to workflow name)
    agent_id: ci-pipeline
    
    # Optional: Files to index (defaults to all markdown files)
    include_patterns: |
      docs/**/*.md
      README.md
      CHANGELOG.md
    
    # Optional: Files to exclude
    exclude_patterns: |
      node_modules/**
      .github/**

Use Cases

1. Index Documentation Automatically

Keep your documentation searchable and up-to-date:

name: Sync Docs to MemVault
on:
  push:
    paths:
      - 'docs/**'
      - '*.md'
    branches: [ main ]
jobs:
  sync-docs:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: Index Documentation
        uses: jakops88-hub/long-term-memory-api/.github/actions/memvault-sync@v1
        with:
          memvault_api_key: ${{ secrets.MEMVAULT_API_KEY }}
          include_patterns: |
            docs/**/*.md
            README.md

2. Store Deployment History

Track every deployment with context:

name: Deploy to Production
on:
  push:
    branches: [ main ]
jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      
      - name: Deploy Application
        run: ./deploy.sh
        
      - name: Store Deployment Memory
        if: success()
        uses: jakops88-hub/long-term-memory-api/.github/actions/memvault-sync@v1
        with:
          memvault_api_key: ${{ secrets.MEMVAULT_API_KEY }}
          user_id: production-deployments
          custom_text: |
            Deployed commit ${{ github.sha }} to production.
            Branch: ${{ github.ref_name }}
            Author: ${{ github.actor }}
            Time: ${{ github.event.head_commit.timestamp }}
            Status: Success

3. Learn from Test Failures

Store test failure information to identify patterns:

name: Run Tests
on:
  pull_request:
jobs:
  test:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      
      - name: Run Test Suite
        id: tests
        run: npm test
        continue-on-error: true
        
      - name: Store Test Failure
        if: failure()
        uses: jakops88-hub/long-term-memory-api/.github/actions/memvault-sync@v1
        with:
          memvault_api_key: ${{ secrets.MEMVAULT_API_KEY }}
          user_id: test-failures
          custom_text: |
            Test suite failed in PR #${{ github.event.pull_request.number }}
            Commit: ${{ github.sha }}
            Author: ${{ github.actor }}
            Files changed: ${{ github.event.pull_request.changed_files }}

4. Context-Aware Code Reviews

Retrieve relevant past discussions before commenting:

name: PR Context
on:
  pull_request:
    types: [opened]
jobs:
  add-context:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      
      - name: Get Similar PRs
        id: context
        uses: jakops88-hub/long-term-memory-api/.github/actions/memvault-retrieve@v1
        with:
          memvault_api_key: ${{ secrets.MEMVAULT_API_KEY }}
          query: ${{ github.event.pull_request.title }}
          limit: 3
          
      - name: Comment on PR
        uses: actions/github-script@v6
        with:
          script: |
            const memories = ${{ steps.context.outputs.memories }}
            if (memories.length > 0) {
              github.rest.issues.createComment({
                issue_number: context.issue.number,
                owner: context.repo.owner,
                repo: context.repo.repo,
                body: `### Similar Past PRs\n\n${memories.map(m => `- ${m.text}`).join('\n')}`
              })
            }

How It Works

When the action runs:

Scans Repository: Finds all matching files based on include/exclude patterns
Extracts Content: Reads file contents and metadata (path, commit, author)
Stores Memories: Sends each file to MemVault's /api/memory/add endpoint
Metadata Enrichment: Automatically adds Git context (commit SHA, branch, author)

Each stored memory includes:

Text: File content or custom message
Metadata: Git context (commit, branch, author, timestamp)
User ID: Repository identifier
Agent ID: Workflow identifier

This allows you to query memories by repository, workflow, or content similarity.

Advanced: Custom Actions

You can also use the MemVault API directly in GitHub Actions:

- name: Custom Memory Storage
  run: |
    curl -X POST https://moderate-krystal-memvault-af80fe26.koyeb.app/api/memory/add \
      -H "x-api-key: ${{ secrets.MEMVAULT_API_KEY }}" \
      -H "Content-Type: application/json" \
      -d '{
        "userId": "${{ github.repository }}",
        "text": "Custom event: Something important happened",
        "metadata": {
          "commit": "${{ github.sha }}",
          "workflow": "${{ github.workflow }}",
          "custom_field": "value"
        }
      }'

Best Practices

Use descriptive user IDs: Organize memories by project/environment
Include rich metadata: More context = better retrieval
Don't over-sync: Only sync on relevant events (main branch, releases)
Handle failures gracefully: Use continue-on-error: true for non-critical syncs
Monitor your usage: Check your dashboard to track API calls and costs

Troubleshooting

Action Fails with "Invalid API Key"

Verify secret name is MEMVAULT_API_KEY (exact match)
Check the key hasn't been revoked in your dashboard
Ensure you're using the correct secret reference: ${{ secrets.MEMVAULT_API_KEY }}