Agents

Tool-Use Critic

Audit an agent trace for wasted calls and loops.

The Prompt

Here is an agent's execution trace:
[paste trace]

Audit it:
1. Which tool calls were unnecessary (the answer was already available)?
2. Where did the agent loop or retry without new information?
3. Where did it guess a parameter instead of asking or looking it up?
4. Rewrite the system prompt section that would prevent the worst failure.

 Fill in before using

[trace]

How to use it

→ Paste the prompt into your model of choice — it works with any frontier model.
→ Replace every bracketed placeholder with your specifics. Concrete inputs beat vague ones every time.
→ If the model asks a clarifying question, answer it — these prompts are designed to invite that.
→ Iterate on the output, not the prompt. The structure is already tuned.

agent debuggingevals

More Agents prompts

Agent System Prompt Skeleton

A battle-tested structure for tool-using agents.

Human Handoff Policy

When the agent should stop and escalate.

Multi-Agent Task Decomposer

Split a job across agents without overlap or gaps.