Overview Benchmark Contents Activity Versions 1 Ratings 0
What's inside The instructions your model reads when the skill loads.
SKILL.md 144 lines 2 files
Feature Refiner
Transform rough feature ideas into clear, actionable feature definitions optimized for continuous delivery.
Process Overview
Follow this collaborative refinement process:
Understand the raw input - Parse unstructured notes and identify all distinct ideas
Identify the core feature - Determine the primary user problem or capability
Organize related points - Group notes that belong to the same feature
Surface separated concerns - Flag notes that should become separate features
Apply continuous delivery principles - Ensure the feature is appropriately sized
Define minimum ideal version - Identify the smallest valuable implementation
Produce clean specification - Create structured markdown output
Step 1: Understand the Input
Parse the user's unstructured notes to:
List all distinct ideas or points mentioned
Identify any explicit requirements or constraints
Note any technical details or dependencies mentioned
Flag any ambiguous or unclear points
Step 2: Identify the Core Feature
Determine what the main feature is by asking:
What user problem is being solved?
What is the primary capability being added?
If multiple capabilities are mentioned, which is the central one?
Confirm with the user : "Based on your notes, it seems the core feature is [X]. Is that correct?"
Step 3: Organize Related Points
Group the notes into:
Core feature points : Directly related to the main capability
When it activates Automatically, when a task matches.
Use when the user asks to refine, clarify, break down, or scope feature ideas for their applications or projects. Common triggers include "help me refine this feature", "break down this idea", "scope this feature", or when the user provides unstructured notes about what they want to build.
Track record Benchmark 10 / 22
test cases passed, vs 4 / 22 without
Adoption
Users 0
Times used 0 View activity → Safety and source details
SkillSafety ✓ 3 of 3 passed ✓ Code scan 15 of 15 checks clean
0 critical · 0 warning · 0 info — scanner v4 — 8/1/2026
View 15 scanner checks live_secret remote_code_execution reverse_shell data_exfiltration ssrf_metadata agent_snooping anti_refusal destructive_commands obfuscated_unicode unicode_homoglyph opaque_blob suspicious_url tool_overreach trigger_abuse prompt_injection_phrasing
✓ Intent Safe
LLM judge · does the skill do what it claims
✓ Content Clean
LLM judge · policy & harmful-content review
Source Imported weareoxd / design-skill-generator
.claude/skills/refine-feature/SKILL.md
View on GitHub ↗ License MITImported 8/1/2026Version v1Bundled 1 file