Inspect AI
Researchinspect-aiInspect AI is an Agent Skill for research published in the a2ui-project/a2ui repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.
Useful when working with the Inspect AI evaluation framework (e.g., in the eval/ directory). Provides instructions on how to access Inspect AI documentation.
How to use this skill
Bring this guide into your coding agent with a prompt tailored to the tool you use.
- Open your project in Codex.
- Copy and paste the prompt below.
- Review files and risks before approving any action.
Inspect the Agent Skill at https://github.com/a2ui-project/a2ui/blob/HEAD/.agents/skills/inspect-ai/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/inspect-ai, but wait for my approval before writing files or running scripts.
Copying this prompt does not install or run the skill. Review third-party files before use.
Inspect AI Framework Guidelines
When you are tasked with creating, modifying, or debugging evaluation scripts (especially within the eval/ directory) that utilize the Inspect AI framework, you should consult the official documentation to ensure you are using the correct APIs, solvers, scorers, and best practices.
Documentation Source
To learn how to use Inspect AI, read the official LLM-optimized documentation guide by fetching the following URL using your read_url_content tool:
https://inspect.aisi.org.uk/llms.txt
Always reference this document when you need clarification on how Inspect AI handles task execution, scoring objects, parallelism, or tool calling architectures.