{"id":45283614,"url":"https://github.com/knid/agentx","last_synced_at":"2026-02-21T02:42:06.529Z","repository":{"id":336961736,"uuid":"1151829048","full_name":"knid/agentx","owner":"knid","description":"Discover, install, and run AI agents powered by Claude. The open marketplace for CLI-first AI agents.","archived":false,"fork":false,"pushed_at":"2026-02-07T01:08:41.000Z","size":279,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-02-07T11:56:23.648Z","etag":null,"topics":["agent","agents","ai-agents"],"latest_commit_sha":null,"homepage":"https://agentx-web.vercel.app","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/knid.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-02-07T00:28:30.000Z","updated_at":"2026-02-07T01:08:44.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/knid/agentx","commit_stats":null,"previous_names":["knid/agentx"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/knid/agentx","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/knid%2Fagentx","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/knid%2Fagentx/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/knid%2Fagentx/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/knid%2Fagentx/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/knid","download_url":"https://codeload.github.com/knid/agentx/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/knid%2Fagentx/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29671825,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-21T00:11:43.526Z","status":"online","status_checked_at":"2026-02-21T02:00:07.432Z","response_time":107,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent","agents","ai-agents"],"created_at":"2026-02-21T02:42:05.580Z","updated_at":"2026-02-21T02:42:06.509Z","avatar_url":"https://github.com/knid.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cp align=\"center\"\u003e\n  \u003cstrong\u003eagentx\u003c/strong\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  Install, chain, and orchestrate AI agents from the terminal.\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"https://www.npmjs.com/package/agentx\"\u003e\u003cimg src=\"https://img.shields.io/npm/v/agentx.svg\" alt=\"npm version\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://opensource.org/licenses/MIT\"\u003e\u003cimg src=\"https://img.shields.io/badge/license-MIT-blue.svg\" alt=\"License: MIT\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://nodejs.org\"\u003e\u003cimg src=\"https://img.shields.io/badge/node-%3E%3D18-green.svg\" alt=\"Node.js \u003e= 18\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\n---\n\n**agentx** lets you discover, install, chain, and orchestrate AI agents from the terminal. Pipe one agent's output into another with standard Unix pipes — research, analyze, write, and ship in a single command. Agents are reusable packages for [Claude Code](https://docs.anthropic.com/en/docs/claude-code) that bundle system prompts, MCP servers, and secrets.\n\n\u003c!-- TODO: Replace with actual demo GIF --\u003e\n\u003c!-- ![agentx demo](./docs/assets/demo.gif) --\u003e\n\n## Features\n\n- **Chain agents with pipes** - `agentx run agent-a --quiet | agentx run agent-b \"use this\"`\n- **Run agents** - Execute agents locally with `agentx run \u003cagent\u003e \"prompt\"`\n- **Install from registry** - One command install: `agentx install @scope/agent`\n- **Schedule agents** - Cron-based scheduling: `agentx schedule start \u003cagent\u003e`\n- **Search \u0026 discover** - Find agents via CLI or browse [agentx.dev](https://agentx.dev)\n- **Publish agents** - Share your agents with `agentx publish`\n- **Scaffold agents** - Create new agents with `agentx init`\n- **Encrypted secrets** - AES-256-GCM encrypted secrets per agent\n- **MCP integration** - Agents declare MCP servers for tool access\n\n## Prerequisites\n\n- **Node.js** \u003e= 18\n- **Claude CLI** installed and authenticated (`npm install -g @anthropic-ai/claude-code`)\n\n## Install\n\n```bash\nnpm install -g @knid/agentx\n```\n\nVerify your setup:\n\n```bash\nagentx doctor\n```\n\n## Quick Start\n\n```bash\n# Search for an agent\nagentx search \"data analysis\"\n\n# Install it\nagentx install @agentx/data-analyst\n\n# Run it\nagentx run data-analyst \"analyze trends in this data\" --file sales.csv\n\n# Chain agents — research, then write\nagentx run web-researcher --quiet \"2026 AI trends\" \\\n  | agentx run writing-assistant \"turn this into a blog post\"\n\n# Interactive mode\nagentx run data-analyst -i\n```\n\n## Commands\n\n| Command | Description |\n|---------|-------------|\n| `agentx run \u003cagent\u003e [prompt]` | Run an agent with a prompt |\n| `agentx install \u003cagent\u003e` | Install an agent from the registry |\n| `agentx uninstall \u003cagent\u003e` | Remove an installed agent |\n| `agentx update [agent\\|--all]` | Update agents to latest versions |\n| `agentx list` | List installed agents |\n| `agentx search \u003cquery\u003e` | Search the agent registry |\n| `agentx trending` | Show trending agents |\n| `agentx info \u003cagent\u003e` | Show agent details |\n| `agentx init` | Scaffold a new agent project |\n| `agentx validate` | Validate an agent manifest |\n| `agentx test` | Test an agent locally |\n| `agentx publish` | Publish an agent to the registry |\n| `agentx schedule start \u003cagent\u003e` | Start an agent's cron schedule |\n| `agentx schedule stop \u003cagent\u003e` | Stop an agent's schedule |\n| `agentx schedule list` | List all active schedules |\n| `agentx schedule logs \u003cagent\u003e` | View execution logs for a scheduled agent |\n| `agentx schedule resume` | Resume all schedules after restart |\n| `agentx configure \u003cagent\u003e` | Configure secrets for an agent |\n| `agentx login` | Authenticate with GitHub |\n| `agentx logout` | Clear authentication |\n| `agentx whoami` | Show current user |\n| `agentx doctor` | Check system requirements |\n| `agentx config` | Manage global configuration |\n\n## Create Your Own Agent\n\n```bash\n# Scaffold a new agent\nagentx init\n\n# Edit the generated files\n# - agent.yaml    (manifest: name, version, MCP servers, permissions)\n# - system-prompt.md  (the system prompt for Claude)\n\n# Validate it\nagentx validate\n\n# Test locally\nagentx run . \"test prompt\"\n\n# Publish to the registry\nagentx login\nagentx publish\n```\n\n### Agent Manifest (agent.yaml)\n\n```yaml\nname: my-agent\nversion: 1.0.0\ndescription: A helpful agent\nauthor: \"@yourusername\"\ncategory: productivity\npermissions:\n  filesystem: true\n  network: true\npre_run:\n  - command: my-bridge\n    background: true\nmcp_servers:\n  filesystem:\n    command: npx\n    args: [\"-y\", \"@modelcontextprotocol/server-filesystem\", \"./\"]\nsecrets:\n  - name: API_KEY\n    description: API key for the service\n    required: true\nschedule:\n  - name: \"Daily report\"\n    cron: \"0 9 * * 1-5\"\n    prompt: \"Generate the daily report\"\n```\n\n## Agent Chaining\n\nAgents write to stdout, so you can chain them with standard Unix pipes. The output of one agent becomes the input context for the next — build multi-step AI workflows in a single line:\n\n```bash\n# Research a topic, then create a Notion page from the results\nagentx run web-researcher --quiet \"2026 AI trends\" \\\n  | agentx run notion-agent \"create a new page summarizing this research\"\n\n# Scan for vulnerabilities, then create a Linear issue for each finding\nagentx run security-scanner --quiet \"audit src/auth/ for vulnerabilities\" \\\n  | agentx run linear-agent \"create a bug for each critical finding\"\n\n# Analyze data, then draft a report\nagentx run data-analyst --quiet \"summarize quarterly revenue\" --file q4.csv \\\n  | agentx run writing-assistant \"turn this into an executive summary\"\n\n# Review code, then send the review to Slack\nagentx run code-reviewer --quiet \"review the latest changes in src/api/\" \\\n  | agentx run slack-agent \"post this code review summary to #engineering\"\n\n# Three-step pipeline: research → rewrite → post\nagentx run web-researcher --quiet \"latest React best practices 2026\" \\\n  | agentx run writing-assistant --quiet \"rewrite as a concise team guide\" \\\n  | agentx run slack-agent \"post this to #frontend\"\n```\n\nUse `--quiet` on intermediate agents to suppress headers/footers and pipe only the raw output. The last agent in the chain can run without `--quiet` to display formatted output.\n\n## Scheduling\n\nAgents can declare cron-based schedules in `agent.yaml`. A shared background daemon runs on your machine and executes agents at the specified times.\n\n```bash\n# Start an agent's schedule\nagentx schedule start slack-agent\n\n# View active schedules\nagentx schedule list\n\n# Check execution logs\nagentx schedule logs slack-agent\n\n# View all past runs\nagentx schedule logs slack-agent --all\n\n# Stop a schedule\nagentx schedule stop slack-agent\n\n# Resume all schedules after a restart\nagentx schedule resume\n```\n\nThe daemon automatically retries failed runs (up to 2 retries with backoff), rotates logs (keeps last 50 per agent), and cleans up when all schedules are stopped.\n\n## Official Starter Agents\n\n14 agents across all 10 categories — install any with `agentx install @agentx/\u003cname\u003e`:\n\n| Agent | Category | Description |\n|-------|----------|-------------|\n| `@agentx/gmail-agent` | communication | Email assistant with Gmail MCP |\n| `@agentx/slack-agent` | communication | Messaging assistant with Slack MCP |\n| `@agentx/whatsapp-agent` | communication | WhatsApp messaging via local bridge |\n| `@agentx/github-agent` | devtools | PR and issue management with GitHub MCP |\n| `@agentx/code-reviewer` | devtools | Code review with GitHub + filesystem MCP |\n| `@agentx/data-analyst` | data | CSV/JSON data analysis with filesystem MCP |\n| `@agentx/postgres-agent` | data | PostgreSQL query and schema explorer |\n| `@agentx/web-researcher` | research | Web search and synthesis with Brave + Fetch |\n| `@agentx/notion-agent` | productivity | Notion workspace and database management |\n| `@agentx/linear-agent` | productivity | Linear issue tracking and sprint management |\n| `@agentx/sentry-agent` | monitoring | Error triage and stack trace analysis |\n| `@agentx/puppeteer-agent` | automation | Browser automation, screenshots, and scraping |\n| `@agentx/writing-assistant` | writing | Proofreading, drafting, and document editing |\n| `@agentx/security-scanner` | security | Vulnerability scanning and dependency auditing |\n\n## Configuration\n\nGlobal config is stored at `~/.agentx/config.yaml`:\n\n```bash\n# View all config\nagentx config list\n\n# Get a value\nagentx config get registry\n\n# Set a value\nagentx config set telemetry false\n```\n\n| Key | Default | Description |\n|-----|---------|-------------|\n| `registry` | `https://registry.agentx.dev` | Registry URL |\n| `claude_path` | `claude` | Path to Claude CLI |\n| `default_output` | `text` | Default output format (text/json) |\n| `telemetry` | `true` | Enable anonymous telemetry |\n| `auto_update` | `true` | Auto-check for updates |\n\n## Project Structure\n\n```\npackages/\n  cli/       # agentx CLI (npm package)\n  web/       # agentx.dev website and registry API\n  agents/    # Official starter agents\n```\n\n## Development\n\n```bash\n# Clone the repo\ngit clone https://github.com/agentx-dev/agentx.git\ncd agentx\n\n# Install dependencies\nnpm install\n\n# Build the CLI\nnpm run build --workspace=packages/cli\n\n# Run tests\nnpm test --workspace=packages/cli\n\n# Type check\nnpx tsc --noEmit --project packages/cli/tsconfig.json\n\n# Link for local development\ncd packages/cli \u0026\u0026 npm link\n```\n\nSee [CONTRIBUTING.md](./CONTRIBUTING.md) for full development guidelines.\n\n## License\n\n[MIT](./LICENSE)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fknid%2Fagentx","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fknid%2Fagentx","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fknid%2Fagentx/lists"}