- Home
- Skills
- DevOps & Cloud
- Ansible Playbook Controller
Works with the AI Agents you already use
Ansible Playbook Controller
Run Ansible playbooks through an MCP server and follow every run: dry-run first, live logs, history and artifacts. One-command install via uvx — ansible-core included.
Free
Ansible Playbook Controller
Example session with this skill installed
Deploy the playbook at deploy-web.yml to the lab hosts, but show me first what it would change. If anything fails, tell me exactly why.
- Read your context and instructions
- Compiled the ansible playbook controller
| Host | Would change |
|---|---|
| web1.lab | nginx.conf, 1 package |
| web2.lab | nginx.conf, 1 package |
| web3.lab | — (already in the desired state) |
Live run
Run 42f1 finished: success, 2 changed, 1 ok, duration 41 s. The playbook text and inventory of this run are snapshotted, so it can be compared with any later run.
Next steps
get_task_logs(42f1) shows the full per-task output if you want it
Re-run any time: run_playbook(playbook_name="deploy-web", provider="lab")
Connects securely to your tools. The creator never sees your data.
About this skill
Gives your agent a safe, recorded way to run Ansible. The agent hands over a playbook and an inventory; the controller executes it, keeps the history, logs and artifacts of every run, and reports back a task id you can follow.
The workflow the skill teaches: syntax-check → dry-run (check=true) → run → follow the logs. A dry run is the only honest answer to "what will this change?", and the skill makes the agent do it before touching your hosts.
What it deliberately is not: it does not write playbooks, lint them, or judge whether running one is wise — the agent decides, the controller executes and records. If you want an assistant that authors Ansible, pair it with a lint/generation skill; this one is the executor with a memory.
Install once: uv tool install ansible-mcp-controller (Python 3.11+, ansible-core arrives with it), or one uvx line in any mcpServers block. Runs from a workstation over stdio, as a container over HTTP, or as a Debian service.
Open source, Apache-2.0: github.com/ppavlovrus/simple_ansible_controller
No support for restarting or scheduling runs. Cannot generate or style-lint playbooks. Credentials must be pre-configured on the server via environment variables.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 11 days ago
- Free with an Agensi account
Requires
Permissions
The agent talks to a local MCP server over stdio; it is the server, not the agent, that opens SSH to the managed hosts. Env vars: ANSIBLE_MCP_DATA_DIR for storage; ANSIBLE_MCP_API_KEY only when the server serves HTTP.