Configure an AIML.chat Agent is a free, MIT-licensed agent skill that sets an agent's type, system prompt, grounding sources, and MCP tools in the right order, so it answers from your content and acts only behind write-confirm. It saves you from a misconfigured agent that guesses or has too much access.
What it does
The skill walks the full configuration: pick the agent type (support, sales, technical, or custom), write or paste a grounded system prompt, attach the content the agent answers from, and connect MCP tools with least-privilege scoping. It enforces the safe defaults — read tools enabled, mutating tools gated behind human approval — so an agent can never quietly change data. It also sets handoff rules when you run more than one agent, so a conversation reaches the right specialist. The result is an agent that's accurate, scoped, and safe from day one.
The skill
name: aiml-configure-agent
Use when setting up or fixing an AIML.chat agent's behavior and tools.
Steps:
1. Choose the agent type that matches the job (support/sales/technical/custom).
2. Set a grounded system prompt: answer only from context, cite sources, escalate.
3. Attach grounding content (URL crawl, repo, files, or MCP source).
4. Connect MCP tools, scope to least privilege, keep writes behind write-confirm.
5. If multiple agents, define handoff rules between them.
How to use it
Copy the SKILL.md into your agent framework's skills folder, or follow it manually in AIML.chat under Agents. It pairs with the builder skills that generate a prompt for you — see the support agent builder — and the prompt reviewer to sharpen what you wrote. The agent overview explains how the types and handoff fit together.
Ready to build one? Start free and configure your first agent in minutes — see pricing for agent and message limits.
Is this skill free?+
Yes — it is MIT-licensed and free to copy. Running the agent counts toward your plan's limits; see pricing.
Does it make writes safe?+
Yes. It scopes MCP tools to least privilege and keeps every mutating tool behind AIML.chat's human write-confirm flow.
Can it set up handoff between agents?+
Yes — when you run more than one agent, the skill defines the rules that route a conversation to the right specialist.