- Home
- Skills
- Softaworks
- Agent Toolkit
- Requirements Clarity
requirements-clarity_skill
- Python
273
GitHub Stars
2
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill softaworks/agent-toolkit --skill requirements-clarity- README.md9.1 KB
- SKILL.md9.1 KB
Overview
This skill clarifies ambiguous product or engineering requirements through focused, iterative dialogue before any implementation begins. It asks two core checks—Why? (YAGNI) and Simpler? (KISS)—and drives the conversation until a high clarity score is reached and a developer-ready PRD can be produced. It prevents rework by turning vague requests into measurable acceptance criteria and phased execution plans.
How this skill works
When invoked, the skill detects vague or incomplete requirements and performs an initial clarity assessment across functional, technical, implementation, and business dimensions. It asks 2–3 targeted questions per round, updates a 0–100 clarity score, and documents each clarification round. Once the score meets the quality threshold (typically ≥ 90), it generates a concrete PRD with acceptance criteria and execution phases ready for handoff.
When to use it
- Requirement descriptions are high-level or vague (e.g., "add login feature").
- New features estimated >2 days of work or involving multiple teams.
- Cross-team integrations or unclear technical constraints are present.
- No acceptance criteria, success metrics, or edge cases specified.
- Before starting implementation to avoid scope creep and rework.
Best practices
- Answer questions incrementally — expect 2–3 clarifying prompts per round.
- Provide concrete examples, tech choices, and priority trade-offs when possible.
- State what is explicitly out of scope to define boundaries.
- Agree on measurable success metrics and acceptance criteria early.
- Iterate until the clarity score is ≥ 90 before generating the final PRD.
Example use cases
- A product manager says "build user management" without roles, workflows, or constraints.
- A request to "implement payments" with no PCI, currency, or gateway details.
- A cross-team feature that requires API contracts and performance targets.
- A complex dashboard feature lacking data sources, refresh cadence, and KPIs.
- A feature that may be low-value and needs a YAGNI evaluation before work.
FAQ
Most clarifications finish in a few quick rounds; complex features may need multiple iterations until the clarity score reaches the threshold.
What happens if users can't answer technical questions?
The skill suggests default options and trade-offs, marks assumptions, and requests confirmation before locking the PRD.