# Agentrun Deploy

> Deploy an agent to AgentRun — Docker-isolated runtime with scheduled (cron) or event-triggered (webhook) process models, syscall mediator, kill switches. PLANNED — not yet implemented.

- Skill: `theagentos/agentrun-deploy` (Agent Skill)
- Install (CLI): `npx skillmds@latest add theagentos/agentrun-deploy`
- Raw SKILL.md: https://api.skillmd.com/api/skills/theagentos/agentrun-deploy/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- Author: theagentos (https://skillmd.com/u/theagentos)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/theagentos/agentrun-deploy

---


# AgentRun deploy — PLANNED

This skill is planned but not yet written. The intended scope:

1. Detect the agent's entrypoint (`agent.py`, `index.ts`, or user-specified).
2. Generate an `agentrun.yaml` deploy descriptor with:
   - Runtime image (matching the language/framework)
   - Trigger model (cron / webhook / manual)
   - Resource limits (CPU, memory, timeout)
   - Syscall policy (allowed tools, network egress rules)
3. Run `agentrun deploy` (or equivalent control-plane API call) and surface
   the deploy URL + first run trace.
4. Verify the first triggered run completes and emits a `process_exit` event.

Track progress at https://github.com/TheAgentOS/agentos-skills/issues