Guide · Agent skill
How to use the Analyze Feature Requests agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The analyze-feature-requests skill is a set of instructions that tells an AI agent how to read a list of customer feature requests, group them into themes, and rank the strongest candidates against your product goals. It is written for product managers, product teams and founders who need to triage a backlog. After this guide you will know what the skill does, how to install it, how to ask for an analysis in plain language, and where its limits are.
Analyze Feature Requestsphuryn/pm-skills · MIT Download the skill Quick start
- 1Download the ZIP. Get the analyze-feature-requests ZIP file from this page.
- 2Unzip into a skills folder. Place the folder in ~/.claude/skills/ for personal use or in a project's .claude/skills/ folder.
- 3Or upload in the Claude apps. Go to Settings, then Capabilities, then Skills, and upload the ZIP.
- 4Prepare your request file. Export your feature requests to a spreadsheet, CSV or document.
- 5State your product goal. Describe the objective and outcomes that should guide the ranking.
- 6Ask for the analysis. Make a natural request to triage or prioritize the backlog and attach the file.
- 7Review and test. Check the themes and top three picks, then test the riskiest assumptions with your team.
What is an agent skill and what does this one do?
An agent skill is a folder containing a SKILL.md file. The file gives an AI agent specific instructions for one kind of task, and the agent loads it when your request matches the skill's description. This skill is from the phuryn/pm-skills collection, under the pm-product-discovery section, and is licensed MIT.
Once installed, the agent groups a list of customer feature requests into themes, assesses how each theme aligns with your stated goals, and picks the top three features to build. For each pick it gives the rationale, alternative solutions, high-risk assumptions, and low-effort ways to test those assumptions.
The skill also pushes the discussion toward customer problems rather than customer-proposed solutions. It uses the Opportunity Score idea from Dan Olsen: Importance × (1 − Satisfaction), normalized to 0–1.
Install the analyze-feature-requests skill
You need the skill's ZIP file, which you download from this page. There are two ways to add it.
Option one is to unzip it into a skills folder. Put it in ~/.claude/skills/ to make it available to you across projects, or in a project's .claude/skills/ folder to limit it to that project. The skill folder should contain SKILL.md.
Option two is to upload the ZIP in the Claude apps. Go to Settings, then Capabilities, then Skills, and upload it there.
- Personal install: ~/.claude/skills/
- Project install: .claude/skills/ inside your project
- Claude apps: Settings → Capabilities → Skills
How do you trigger the skill with a natural request?
You do not need a special command. The skill's description covers reviewing customer feature requests, triaging a backlog and making prioritization decisions, so ask for one of those things in plain language and attach your file. The agent reads the file directly, and it also expects you to describe your product goal.
The skill asks you to confirm the product objective before it ranks anything. If you skip the goal, expect a clarifying question rather than a ranking.
- Here is our feature request export and our goal this quarter is reducing churn for small teams. Triage it and tell me the top three to build.
- Group these customer requests into themes and check how each one fits our goal of growing enterprise accounts.
- I'm making a prioritization decision next week. Analyze this backlog CSV against our onboarding goals and explain the reasoning.
What input does the skill need?
The skill works on files you provide. The source describes spreadsheets, CSVs and documents containing feature requests. If the data is structured, the agent may create a summary table. Output is saved as markdown or in a structured document.
The skill does not connect to Jira, a support desk or any other tool. It only reads what you hand it, so export your requests first.
- Spreadsheets, CSV files or documents with feature requests
- A stated product goal or desired outcomes
- No additional software or packages are listed as required
What to check before you rely on the output
The skill runs no scripts and makes no network calls, according to its listing. Its scoring is judgment-based, so the agent's ranking is a starting point for discussion, not a decision. The skill does not use real usage or revenue data, so results do not reflect how much customers actually use or pay for a feature.
Treat the output as material for a team review. Check the themes against the original requests, and challenge any assumption that looks wrong.
- No scripts run and no network access is used, per the listing
- Scores are judgment-based and need team review
- Real usage or revenue data is not part of the analysis
Tips for better results
Clear inputs produce clearer themes. State your goal in one or two sentences with specific outcomes, and include enough request detail for the agent to see what customers are actually asking for.
Use the output to reframe requests. Because the skill favors problems over proposed solutions, the alternatives it lists are often more useful than the feature as written.
- Write a concrete goal, such as a metric or customer segment
- Include the request text and any context columns in your file
- Ask follow-up questions on any of the three picks to go deeper
What people use it for
- Triaging a backlog of customer feature requests exported from a spreadsheet before a planning meeting
- Preparing a prioritization discussion with stated quarterly goals
- Grouping raw customer feedback from documents into themes for a founder review
- Generating alternative solutions and risky assumptions for a shortlisted feature
Frequently asked questions
How do I install the analyze feature requests skill?
Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. In the Claude apps, you can instead upload the ZIP under Settings, then Capabilities, then Skills.
Is the analyze-feature-requests skill free?
The skill is licensed under MIT, according to its listing. Check the repository's license file for the full terms before using it in your work.
What does the skill need to run?
The listing states that it requires nothing beyond the skill itself. It runs no scripts and uses no network access, and it only reads files you provide.
Can it pull feature requests from Jira or a support tool?
No. The skill only reads files you provide, such as spreadsheets, CSVs or documents. Export requests from your tool first, then attach the file.
Why does the agent ask for my product goal first?
The skill is designed to rank requests against product goals. Without a defined goal, the results would be arbitrary, so the agent asks you to confirm the objective before ranking.
What are alternatives to this skill for feature prioritization?
The skill's listing does not compare it with other tools, so this page cannot make that claim. Its own further reading points to the Kano Model and Continuous Product Discovery material, which cover related approaches.
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 product manager, product team or founder who needs a structured first pass over a backlog of customer requests and has clear product goals. Pick something else if you have no defined goal, need final decisions made automatically, or want the tool to pull data from Jira or a support system, since it does none of those.
Download the skill