Broccoli: Open-source harness for running AI coding agents from Linear tickets in cloud sandboxes
What Broccoli does
Broccoli turns Linear tickets into shipped PRs by running AI coding agents in isolated cloud sandboxes. Each task gets its own sandbox where it checks out the repo, uses ticket context, implements solutions, runs tests and review loops, then opens a pull request for team review.
Architecture and deployment
The implementation uses:
- Webhook deployment: GCP
- Sandbox: GCP or Blaxel
- Project management: Linear
- Code hosting & CI/CD: GitHub
It's deployed to your own GCP project against your Postgres database with no third-party control plane. The deployment is production-grade using Cloud Run + Secret Manager + webhook deduplication + durable job state.
Key features
- Linear ticket → reviewable PR workflow
- Your infrastructure, keys, and data stay within your tenancy
- AI code review on every PR using Claude and Codex
- Customizable prompt templates that you can fork, tune, and version
- Deploys in ~30 minutes with one bootstrap script and one config file
Deployment prerequisites
To deploy, you need:
- GCP project with billing attached and
gcloudlogged in - GitHub App with specific permissions (Contents, Pull requests, Issues = read/write; Metadata = read-only; subscribed to Pull request review event)
- Linear bot user (dedicated, not personal account) with API key
- OpenAI and Anthropic API keys with billing enabled
- Linear webhook pointing at
${Service URL}/webhooks/linearusing auto-generatedbroccoli-oss-linear-webhook-secret
Usage results
In the last four weeks, 100% of PRs from non-developers were shipped via Broccoli. For developers on the team, this share is around 60%, with more complicated features requiring manual back-and-forth design with Codex/Claude Code.
📖 Read the full source: HN AI Agents
👀 See Also
Self-Maintaining Documentation System Using Fenced Blocks for Zero Drift
A developer built a bash script that extracts structured data directly from source files and injects it into CLAUDE.md through fenced HTML comment blocks, ensuring documentation stays in sync with code without manual maintenance.
Strands Decider 2B: A 2B-Parameter Decision Model for Agentic Workflows
Strands Labs releases Strands Decider 2B, a 2B-parameter decision model that replaces the LM head with a pointer head to pick between options in ~115ms on an RTX 3090. Weights and training scripts are on GitHub and Hugging Face.
SIDJUA Framework Adds Governance Layer to Autonomous AI Agents
SIDJUA is a framework with built-in governance, role-based authority rules, and full audit trails that sits on top of any AI model with an API. The demo shows a three-tier hierarchy that scales to 7+1 tiers, with every decision logged and costs tracked in real time.
Engram Memory SDK: Graph-Based Memory for AI Agents with Local Models
Engram Memory SDK is an open-source graph memory system for AI agents that works with local models via LiteLLM. It requires only one LLM call for ingestion, then uses vector search and graph traversal for recall with zero ongoing LLM costs.