name: skill-name description: A clear, complete description of what this skill does and the exact conditions under which an agent should invoke it. Start with the capability, then list trigger phrases. WHEN: "trigger phrase one", "trigger phrase two".
Skill Name
Overview
One or two sentences stating the purpose of this skill and the outcome it produces.
When to Use
- Concrete condition that should activate this skill.
- Another concrete condition.
- Explicitly note conditions where this skill should NOT be used.
Inputs
| Input | Required | Description |
|---|---|---|
field_name | yes | What it is and where it comes from. |
Process
- First concrete step the agent performs.
- Next step, with the decision or computation involved.
- Continue until the skill produces its output.
Outputs
Describe the structured result this skill returns and how downstream skills consume it.
Rationalizations
| Excuse | Rebuttal |
|---|---|
| "This check can be skipped for a tiny order." | State why the step is still required. |
Red Flags
- Signs that the skill is being applied incorrectly or producing unreliable output.
Verification
- Evidence that must exist before the skill is considered complete (for example, a populated output field, a passing test, or a logged rationale).