Guide · Agent skill
How to use the Anitta Challenge Profile agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The Anitta Challenge Profile is an agent skill, a folder containing a SKILL.md instruction file that changes how an AI agent reasons while it works. Once installed, the agent behaves as a rigorous, supportive reviewer: it questions its first answer, separates evidence from interpretation, states its assumptions, and matches the strength of its wording to the quality of the evidence. This guide explains what the skill does, how to install it, how to trigger it, and where it does not fit.
Anitta Challenge Profilegithub/awesome-copilot · MIT Download the skill Quick start
- 1Download the ZIP. Get the skill ZIP from the page you are reading.
- 2Unzip the folder. Extract the skill folder so that SKILL.md is inside it.
- 3Place it in a skills folder. Put the folder in ~/.claude/skills/ for personal use, or .claude/skills/ inside a project.
- 4Or upload in the Claude apps. Go to Settings, then Capabilities, then Skills, and upload the skill.
- 5Ask for a review in plain words. Describe the work and say you want it challenged before it is acted on.
- 6Give decision context. State the decision, the required confidence and the biggest known uncertainty.
What the Anitta skill makes the agent do
The skill is a persona-style instruction file, not a tool that runs code or fetches data. It tells the agent to challenge the first comfortable answer, make assumptions explicit, and keep challenge constructive and specific.
When the skill is active, the agent works through a three-phase review lens: reasoning and logic first, interpretation and narrative second, and rigor checks and counterfactuals third. It also uses a set of rigor questions, such as what decision is being supported, what evidence is missing, and what observation would change its mind.
- Questions its first answer and looks for overgeneralization
- Separates evidence from interpretation
- States assumptions about data quality, causality and stability
- Matches wording such as 'suggests' or 'demonstrates' to how strong the evidence is
- Can hand off to two related profiles, Quinn (execution) and Wiggins (narrative)
Install the Anitta skill from the ZIP
Download the ZIP for the skill from the page you are reading. Unzip it so that the skill folder, which contains SKILL.md, sits in one of these locations: ~/.claude/skills/ for your personal setup, or .claude/skills/ inside a project folder if you want the skill limited to that project.
You can also upload the skill in the Claude apps. Go to Settings, then Capabilities, then Skills, and add the skill there.
- Personal install: place the unzipped folder in ~/.claude/skills/
- Project install: place it in .claude/skills/ inside your project
- Claude apps: Settings → Capabilities → Skills, then upload
Trigger the skill with natural requests
You do not need a special command. Ask for the kind of review you want in plain words, and mention that you want the work challenged before it is acted on. The skill is written to start meaningful tasks by confirming the exact question, the decision that depends on it, the required confidence level, and the biggest known uncertainty.
If the work is high-stakes, say so. The skill's guidance is to bring it in when the cost of being wrong is meaningful.
- 'Pressure-test this recommendation before I send it to the team.'
- 'What assumptions am I making in this churn analysis?'
- 'Check whether my wording overstates what this evidence shows.'
- 'What would a smart skeptic conclude from the same numbers?'
Example: reviewing an analysis draft
Paste a draft conclusion and ask the agent to challenge it. A useful response names the decision the draft supports, lists the assumptions behind it, flags any claim stronger than the evidence allows, and suggests what missing evidence could change the result.
The skill also includes a query authoring standard: when sharing queries, the agent uses fully qualified object names, with cluster and database prefixes, and avoids bare table names in shared drafts. This applies to the Ember working set the file was written for.
- Ask for the exact decision the draft supports
- Ask which assumptions about data quality or causality are unstated
- Ask what observation would change the conclusion
What the skill needs and how it runs
The skill lists no requirements. It does not run scripts and does not use the network. Because it is only an instruction file, it changes how the agent reasons; it does not add domain knowledge or data. The agent still needs the material you want reviewed.
The file is licensed under MIT.
- Requirements: none listed
- Runs scripts: no
- Uses network: no
- Included files: SKILL.md
Tips for getting useful challenge
Give the agent the decision and the stakes up front. Challenge is most useful when it is specific and linked to a decision; the skill itself notes that generic skepticism slows work.
Use it selectively. For quick, low-stakes tasks, extra challenge only adds delay. Tell the agent when you want a lighter touch and it can adapt.
- State the decision and required confidence before asking for review
- Share the actual evidence, not only conclusions
- Skip the skill for quick, low-stakes work
What people use it for
- Pressure-testing a recommendation before a team acts on it
- Checking that a report's wording matches the strength of its evidence
- Finding unstated assumptions in a data or analytics analysis
- Generating counterfactuals, such as what observation would change the conclusion
- Reviewing shared queries for fully qualified object names in the Ember working set
Frequently asked questions
How do I install the Anitta Challenge Profile skill?
Download the ZIP from the page, then unzip it so the skill folder containing SKILL.md sits in ~/.claude/skills/ or in .claude/skills/ inside a project. You can also upload it in the Claude apps under Settings, then Capabilities, then Skills.
Does the Anitta skill require anything else to run?
The skill lists no requirements. It runs no scripts and uses no network. It is an instruction file that changes how the agent reasons.
Is the Anitta skill free to use?
The file is licensed under MIT. The source does not mention pricing, so check the licence terms on the directory page for any other conditions.
Will the skill give my agent new data or domain knowledge?
No. The skill only changes how the agent reasons. It does not supply data or domain expertise, so the agent still needs the material you want reviewed.
Can I use this skill outside the Ember team?
The skill was written for the Ember collaboration. The query naming rule and the handoffs to Quinn and Wiggins refer to a specific working set, so parts may not apply to other teams. The core challenge habits still work on their own.
Are there alternatives if this skill does not fit my work?
If you only need a quick check, you can ask your agent directly to list its assumptions and the evidence behind each claim. For a general-purpose review workflow, look for other skills in the directory that match your task. The source does not compare this skill with specific alternatives.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are a team or individual who needs an AI reviewer to pressure-test analysis or recommendations where overconfident claims would be costly, especially data or reporting work. Pick something else for quick, low-stakes tasks, for work outside the Ember team's context, or if you expect the skill to bring data or domain knowledge with it.
Download the skill