Build autonomous agents that fail safely.
What it does
Guides the design and implementation of reliable autonomous agents with planning, self-correction, and safety boundaries.
When to use it
Use when the user mentions autonomous agents, goal decomposition, agent loops, reflection, or agent planning.
How to use it
It changes how the agent builds autonomous-agent systems by prioritizing reliability, constrained scope, guardrails, human review, and recoverable failures.
What you provide
- An existing project
What you need · 6
Routes multi-agent coordination work to this skill; whether it must be installed to complete the workflow is unstated.
Routes agent testing and evaluation work to this skill; whether it must be installed to complete the workflow is unstated.
Routes agent tool design and implementation work to this skill; whether it must be installed to complete the workflow is unstated.
Routes persistent-memory architecture work to this skill; whether it must be installed to complete the workflow is unstated.
Access · 0
No additional actions listed.