SkillAgentSearch skills...

Pixel Agent

Autonomous agent swarm platform with governance, budgeting, task orchestration, and observability

Install / Use

/learn @RemyLoveLogicAI/Pixel Agent
About this skill

Quality Score

0/100

Supported Platforms

Universal

README

Pixel-Agent

Autonomous agent swarm platform with governance, budgeting, and observability

CI License: MIT Node: 20+

Pixel-Agent is a full-stack autonomous agent swarm platform that provides the infrastructure for running, coordinating, and governing fleets of AI agents. It handles the full agent lifecycle — from spawning and task assignment to budget enforcement and real-time observability — so you can focus on agent logic rather than plumbing.


Features

  • Agent lifecycle management — spawn, monitor, and terminate agents via a clean REST API; backed by an agent pool with circuit-breaker fault tolerance
  • Budget and governance engine — capability tokens enforce spending limits and access controls; the governance service audits all agent actions
  • Task orchestration and swarm coordination — the swarm engine assigns tasks to agent pools, routes messages via swarmMessageBus, and synthesises results through synthesisService
  • Real-time observability — event traces capture every agent action; /events and /traces endpoints stream structured logs for dashboards or alerting
  • HyperSwarm P2P networking — the CLI spins up a HyperSwarm node so distributed agent clusters can discover and communicate peer-to-peer

Architecture

graph TD
    A[API Server<br/>Express / Hono] --> B[Routes]
    B --> B1[agents]
    B --> B2[tasks]
    B --> B3[budget]
    B --> B4[governance]
    B --> B5[swarms]
    B --> B6[events & traces]
    B --> B7[goals / memory / tools]

    A --> C[Services]
    C --> C1[agentExecutor]
    C --> C2[agentPool]
    C --> C3[capabilityTokenService]
    C --> C4[governanceService]
    C --> C5[swarmEngine + swarmMessageBus]
    C --> C6[synthesisService]
    C --> C7[heartbeatRunner + circuitBreaker]

    A --> D[Middleware]
    D --> D1[error-handler]

    A --> E[CLI]
    E --> E1[hyperSwarm P2P node]

Tech Stack

| Layer | Technology | |---|---| | Runtime | Bun | | API framework | Express 5 / Hono | | P2P networking | HyperSwarm | | ORM | Drizzle ORM | | Language | TypeScript 5 | | Monorepo | pnpm workspaces | | Linting/Formatting | Biome |


Quickstart

Prerequisites

  • pnpm ≥ 9
  • Node.js ≥ 20 (or Bun ≥ 1.1)

1. Clone and install

git clone https://github.com/RemyLoveLogicAI/Pixel-Agent.git
cd Pixel-Agent
pnpm install

2. Configure environment

cp .env.example .env
# Edit .env with your DATABASE_URL and other required values

3. Start the API server

pnpm --filter @workspace/api-server run dev

4. (Optional) Start a HyperSwarm P2P node

pnpm run hyper-swarm

Environment Variables

See .env.example for the full list. Key variables:

| Variable | Description | Default | |---|---|---| | NODE_ENV | Runtime environment | development | | PORT | HTTP server port | 3000 | | DATABASE_URL | Postgres or SQLite connection string | — | | CLOUDFLARE_ACCOUNT_ID | Cloudflare account (optional, for Workers deploy) | — | | CLOUDFLARE_API_TOKEN | Cloudflare API token with Workers:Edit permission | — | | LOG_LEVEL | Pino log level (trace|debug|info|warn|error) | info |


Docker

docker build -t pixel-agent .
docker run -p 3000:3000 --env-file .env pixel-agent

Project Structure

Pixel-Agent/
├── artifacts/
│   ├── api-server/          # Core API — routes, services, middleware, CLI
│   └── mockup-sandbox/      # UI mockup / sandbox environment
├── lib/                     # Shared libraries (db, zod schemas, integrations)
├── scripts/                 # Build and utility scripts
├── .github/workflows/       # GitHub Actions CI
├── biome.json               # Linting and formatting config
├── Dockerfile               # Multi-stage Docker build
├── pnpm-workspace.yaml      # Monorepo workspace definition
└── tsconfig.base.json       # Shared TypeScript config

Contributing

  1. Fork the repo and create a feature branch: git checkout -b feat/your-feature
  2. Make your changes and run pnpm run typecheck to verify types
  3. Open a pull request against develop

Please follow the existing code style — Biome enforces formatting on CI.


Security

See SECURITY.md for our vulnerability disclosure policy. To report a security issue, email security@lovelogicai.com — do not open a public issue.


License

MIT © 2026 LoveLogicAI LLC


Part of the LoveLogicAI Agent Company OS

View on GitHub
GitHub Stars0
CategoryDevelopment
Updated3h ago
Forks0

Languages

TypeScript

Security Score

85/100

Audited on Mar 26, 2026

No findings