Integration Tutorial April 13, 2026

Turn GitHub Issues into Searchable Knowledge

GitHub issues contain the real history of your project -- bug reports, feature discussions, design decisions buried in comment threads. REM Labs indexes all of it so your AI can search across issues, discussions, and PR reviews by meaning instead of by keyword.

What Gets Imported

The GitHub integration extracts data from your repositories and stores each issue and discussion as a richly indexed memory:

Issues -- title, body (markdown), and all comments with author and timestamp
Pull request reviews -- review comments, inline code comments, and review decisions (approved, changes requested)
Discussions -- GitHub Discussions threads with all replies and reactions
Labels and milestones -- stored as filterable metadata on each memory
Cross-references -- issue mentions (#123) and PR links become knowledge graph edges

Connect via GitHub App

In the REM Labs Console, go to Integrations and click Connect GitHub. Install the REM Labs GitHub App on your organization or personal account. Select which repositories to sync.

Import via API

curl -X POST https://api.remlabs.ai/v1/memory/sync/github/import \
  -H "Authorization: Bearer sk-slop-..." \
  -H "Content-Type: application/json" \
  -d '{
  "github_token": "ghp_...",
  "repos": ["myorg/api-server", "myorg/frontend", "myorg/docs"],
  "include_issues": true,
  "include_prs": true,
  "include_discussions": true,
  "state": "all",
  "sync_mode": "incremental"
}'

Setting state: "all" imports both open and closed issues. Use "open" to limit to active issues only. Incremental mode uses the updated_at timestamp to process only issues modified since the last sync.

Webhook for Real-Time Sync

The GitHub App supports webhook events so new issues, comments, and PR reviews are synced to REM Labs in real time. No manual re-sync needed -- when someone opens an issue or leaves a review comment, it enters your knowledge base within seconds.

Search Across Repositories

curl -X POST https://api.remlabs.ai/v1/integrations/github/search \
  -H "Authorization: Bearer sk-slop-..." \
  -H "Content-Type: application/json" \
  -d '{
  "query": "how to handle WebSocket reconnection gracefully",
  "filters": {"repo": "myorg/api-server"},
  "limit": 10
}'

This finds the closed issue "Connection drops under load" and the PR review thread where your team debated exponential backoff strategies -- because semantic search connects "graceful reconnection" with "exponential backoff" and "connection drops" even when those exact words are not in your query.

Cross-References as Graph Edges

When issue #42 mentions "see #38 for the original design," REM Labs creates a graph edge between them. When you search for the topic of issue #42, issue #38 appears as related context through graph traversal -- even if its content is not semantically close to your query. This mirrors how developers naturally navigate issue threads.

API Endpoints Reference

Endpoint	Method	Description
`/v1/memory/sync/github/import`	POST	Import issues, PRs, and discussions from repos
`/v1/integrations/github/status`	GET	Check sync status per repository
`/v1/integrations/github/search`	POST	Search within the GitHub namespace

Rate limit handling: The integration respects GitHub's API rate limits and uses conditional requests (If-Modified-Since) to minimize quota usage. For large repos (10,000+ issues), initial sync runs as a background job. See GitHub integration docs.

Make your GitHub history AI-searchable

Free tier. Semantic search across issues, PRs, and discussions.

Get started free →