- Home
- Skills
- Business & Operations
- prd assistant
prd assistant
Professional PRD assistant that drafts structured requirements and interactive HTML prototypes from messy inputs.
Free
Works with the AI Agents you already use
prd assistant
Example session with this skill installed
Draft a Type B PRD for a new 'Batch Refund' feature. I've attached a screenshot of the current single refund page. It needs to support CSV uploads and multi-select from the order list.
- Read your context and instructions
- Compiled the prd assistant
I have analyzed the current refund interface and your requirements for the Batch Refund feature. This is classified as a Type B (Module Level) change. I have drafted the PRD focusing on the new upload logic and selection states.
PRD: Batch Refund Module
| Attribute | Detail |
|---|---|
| Complexity | Type B (Module Level) |
| Status | Draft (Pending confirmation on max batch size) |
| Scope | Order Management System (Web) |
1. Functional Overview
Enable administrators to process multiple refunds simultaneously via manual selection or CSV file upload. This extends the existing Single Refund logic.
2. UI & Interaction
Multi-select: Add checkboxes to the Order List table. A sticky bottom bar appears when >0 items are selected, showing "Process Batch Refund (N)".
- Upload Entry: A "Batch Upload" button next to the search bar triggers a modal for CSV drops.
- Validation: Real-time feedback on CSV rows. Invalid Order IDs must be listed in an error callout.
3. Business Rules
- Eligibility: Only orders with status
PaidorPartially Refundedare selectable.
Calculation: Total refund amount defaults to the sum of all refundable totals. Individual overrides are not supported in the batch view (must use single refund for partials).
- Permissions: Restricted to
Finance_LeadandSuper_Adminroles.
4. Data Fields
upload_id: System-generated unique ID for the batch.order_id_list: Array of targeted orders.refund_reason_code: Required global reason for the entire batch.
Next steps
- Confirm the maximum number of orders allowed in a single CSV upload (currently assumed 500).
- Review the error handling for partially failed batches (e.g., one order fails, do others proceed?).
- Approve the drafted logic to begin the HTML prototype generation for the upload modal.
Connects securely to your tools. The creator never sees your data.
About this skill
Stop wasting hours on bloated PRDs that developers ignore and designers misunderstand. This skill converts messy meeting notes, chat logs, and rough screenshots into structured, implementation-ready product requirement documents and interactive HTML prototypes. It filters out fluff to focus on logic, state transitions, and business rules, ensuring your technical team has exactly what they need to build without the overhead of project management jargon.
What it does
- Multi-mode processing lets you draft new PRDs from scratch, modify existing local files, or perform a strictly critical audit without altering the source.
- Complexity-aware drafting scales the detail level from simple UI tweaks (Type A) to high-level module changes (Type B) and complex, multi-role business logic (Type C).
- Interactive prototyping generates standalone HTML prototypes automatically when you provide UI screenshots or explicit layout requirements.
- Requirement memory maintains a persistent
memory.mdfile to track decisions, pending questions, and changes across multiple sessions or handoffs. - Conflict resolution identifies contradictions between legacy rules and new instructions, forcing a decision rather than guessing the outcome.
How it works
- Intake your raw materials like screenshots, chat history, or rough drafts and specify if you need a PRD, a prototype, or both.
- Analysis categorizes the task by complexity (A/B/C) and identifies missing constraints or business logic gaps that require user confirmation.
- Execution produces the structured Markdown PRD following strict language guidelines or generates the HTML prototype assets.
- Validation runs a boundary scan to check for missing edge cases or state transitions before final delivery.
Frameworks & tools
Built for standard Markdown environments and browser-based prototyping. Optional validation scripts require Python 3.10+. It generates vanilla HTML/CSS prototypes for maximum portability across web and mobile previews.
Why this beats prompting it yourself
Generic prompts often invent missing logic or delete boring but critical business constraints. This skill follows a strict "unknowns and conflicts" protocol that halts execution on core logic gaps rather than hallucinating requirements, ensuring your documentation reflects actual product decisions.
Use cases
- Converting a whiteboard photo and voice memo into a Type B module PRD.
- Auditing a legacy PRD against a new set of business constraints to find contradictions.
- Generating a clickable HTML prototype from a "current state" screenshot to demo a proposed UI overhaul.
- Maintaining a long-running requirement log for complex features that span multiple weeks of discussion.
Known limitations
Does not generate technical architecture, API schemas, or deployment plans. Prototyping is limited to static HTML/CSS and basic interactivity; it does not connect to live databases or backends.
How to use
Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scanned
Verified clean 1 day ago
- Free to download with an account
Requires