Requirements Discovery

    1

    It guides your AI agent to clarify the problem, examine competing explanations, separate evidence from assumptions, and choose a practical MVP scope.

    $1

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Requirements Discovery

    Example session with this skill installed

    Use $requirements-discovery to assess this fictional product idea: a lightweight study planner for university students balancing coursework and part-time work.

    I am a solo developer with three weeks for an initial prototype. Assume the current alternatives are a calendar, a to-do app, and class group chat. I have no interviews or usage data. Do not assume access to university systems.

    Clarify the problem and key assumptions. Explore three clearly labeled synthetic scenarios, including a reason someone might not adopt the product. Propose a small MVP with testable acceptance criteria, explain what to defer, and suggest a low-cost validation plan that could change the recommendation.

    Respond in English, within 500 words, in this conversation only. Do not browse the web or create project files. Distinguish assumptions from evidence; do not invent user quotes, research results, or adoption statistics.

    • Read your context and instructions
    • Compiled the requirements discovery

    Recommendation: proposed exploratory prototype. Focus on helping students revise a realistic study plan when work shifts change. Whether this saves enough effort to justify maintaining another tool is the biggest unknown.

    Evidence boundary: Your brief establishes the solo developer, three-week limit, assumed alternatives, and lack of university integrations. There are no interviews or usage observations; demand and benefits remain unverified.

    The proposed problem is: after a deadline or shift changes, a student must reconcile calendar commitments, to-do items, and information from class chat to decide what study work still fits. Possible consequences include repeated planning and infeasible commitments, but neither is established.

    Key assumptions

    • H1: Replanning across tools causes meaningful friction.
    • H2: Students will manually enter and maintain enough information.

    H3: Better visibility changes decisions. A competing explanation is insufficient time or difficulty starting work, which a planner may not resolve.

    The following three synthetic scenarios are model-generated explorations, not interviews or research findings. All behavioral details are invented to identify questions; no real-user calibration has occurred.

    Synthetic scenario — changed shift. A student wants to finish an assignment before Friday. An added Thursday shift displaces study time recorded in their calendar; remaining tasks sit in a to-do app. A combined view might help them relocate work. It fails if updating it duplicates too much effort. Validate by observing a recent real shift change and the steps taken to recover.

    Synthetic scenario — impossible week. A student has several deadlines and extra paid shifts they cannot decline. Their existing tools already show the commitments. The underlying problem may be insufficient capacity. Showing unallocated study hours could support prioritization, but cannot create time. Check whether recognizing overload changes an actual decision.

    Synthetic scenario — non-adoption. A student with predictable shifts already uses calendar blocks and a short task list successfully. Another planner adds maintenance without improving decisions. Observe their current workflow; retain this as a counterexample rather than assuming every working student needs the product.

    Small MVP, conditional on H1–H3: a single-device weekly planner with manual commitments, assignment deadlines, estimated study hours, and movable study blocks.

    Testable acceptance criteria

    • Create, edit, and delete commitments and assignments; reject end times before start times.
    • Place and move study blocks; flag overlaps and blocks after deadlines. Changing a shift reveals affected blocks.
    • Show estimated hours still unallocated; completing a block updates remaining work.
    • Preserve entries after reopening; provide a clear-all action that removes them.

    Defer automatic scheduling, chat parsing, integrations, accounts, synchronization, collaboration, and gamification: these add dependencies before the core benefit is established.

    Low-cost validation: In week one, recruit five accessible working students, inspect recent planning episodes, and compare a paper prototype with their existing tools. Build in week two only if recurring friction emerges; trial during week three. Predeclare a provisional continuation rule: three participants independently reuse it and demonstrate easier replanning against their baseline. This is a decision rule, not population evidence. If calendars suffice, stop; if entry burden dominates, simplify; if overload dominates, reconsider the product’s purpose.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    What it does

    Requirements Discovery helps solo developers and small teams turn unclear product ideas into actionable software requirements.

    It guides your AI agent to clarify the problem, examine competing explanations, separate evidence from assumptions, and choose a practical MVP scope.

    What you get

    • A clear problem statement, relevant stakeholders, and important constraints.
    • Explicit assumptions, unanswered questions, and alternative explanations.
    • Clearly labeled synthetic scenarios to explore possible obstacles, failure paths, and reasons people might not adopt the product.
    • Proposed requirements, scope trade-offs, and testable acceptance criteria.
    • A low-cost validation plan explaining what to observe and how the findings could change the recommendation.

    For complete project analyses, the skill can produce an analysis.md report and a structured case.json linking evidence, claims, requirements, experiments, decisions, and changes.

    When to use it

    Use it when you are exploring a product idea, deciding what belongs in an MVP, preparing for user research, or revising requirements after receiving new feedback.

    For a small question, it can provide a focused answer without generating a full project dossier.

    Research and synthetic exploration

    When browsing is available, the agent can examine public sources and record their relevance and limitations. Without access to sources or users, it identifies assumptions and proposes ways to test them.

    Synthetic scenarios are model-generated exploration, not interviews or statistical samples. They help generate questions and hypotheses; they do not establish market demand, adoption rates, or willingness to pay.

    Requirements and outputs

    Core analysis runs through your AI agent. No skill-specific API key or paid service is required.

    An optional Python 3.10+ validator checks saved case files for structure, references, and selected evidence-state rules. It uses the Python standard library and does not verify factual truth or analytical quality.

    Getting started

    Ask your agent to use requirements-discovery, then describe your idea, current workflow, available evidence, constraints, and the decision you need to make.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Product Management.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 1 day ago

    Nothing to install. It runs in your AI agent.

    Optional
    Python 3.10+
    Web browsing
    orsearch for public-source research

    Core analysis runs in your AI agent. Python 3.10+ is optional for checking saved case.json files; no extra Python packages or API keys are required by this skill. Web research is optional. Your agent's own access and usage costs still apply.

    Permissions

    Read Files
    Write Files
    Terminal / Shell
    Browser

    File Scopes

    requirements-discovery/**
    requirements-inputs/**
    requirements-analysis/**

    Reads the installed skill and project files chosen by the user. Writes requested reports and case data to the chosen output folder. Terminal access is used for optional local JSON validation; browsing is used for optional public-source research. No third-party account is required.

    Listed1 day ago

    What's inside

    Frequently Asked Questions

    This skill includes the core discovery logic, a Python validation script for data integrity, and templates for `analysis.md` and `case.json` to help you track requirements in your version control system.

    Unlike standard prompts that might offer generic encouragement, this skill is programmed to act as a "devil's advocate," identifying high-risk assumptions and missing evidence that could lead to building features users don't actually need.

    The skill is designed to work with any agent that supports Markdown and JSON output; however, the included validation script requires a local Python environment to verify the structural integrity of your discovery files.

    Yes, it is designed for iterative use. You can feed existing discovery files back into the skill to update your problem model as you gather new user feedback or pivot your MVP scope.

    While the skill uses synthetic exploration to simulate user scenarios and edge cases, it does not perform live web scraping or real-time market research; all evidence must be provided via your project context or existing documents.

    New in Product Management