Feature radar ref

Skill runkids/feature-radar/skills/feature-radar-ref

Record external observations, ecosystem trends, and creative inspiration into .feature-radar/references/. MUST use this skill when the user mentions something interesting from outside their project — other tools, articles, approaches, or trends. Even casual mentions like "I saw a cool thing in X" should trigger this skill. Use when the user: - Says "I saw this cool thing in X's API", "check out how X handles this" - Shares a URL, article, talk, or research finding with relevant insights - Notes an ecosystem trend: new tools, standards, community patterns - Mentions a related project shipping a notable feature - Wants to bookmark external inspiration: "interesting approach", "save this" - Says "add reference", "log observation", "track this project" Do NOT use for internal learnings/patterns — that's feature-radar-learn's job. Do NOT use for prioritizing features — that's feature-radar's job.From its SKILL.md

Install

npx -y skills add runkids/feature-radar --skill feature-radar-ref

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

13 stars13 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.6 KB, 644 tokens by cl100k_base, as published. Nobody here has run it

Add Reference

Record external observations into .feature-radar/references/.

Deep Read

<HARD-GATE> Read and follow `../feature-radar/references/DEEP-READ.md` — complete all 6 steps before proceeding. </HARD-GATE>

Behavioral Directives

<HARD-GATE> Read and follow `../feature-radar/references/DIRECTIVES.md`. </HARD-GATE>

Workflow

Identify the source — ask the user what they observed:
- Interesting project, technique, or creative approach?
- Ecosystem trend or emerging pattern?
- Notable feature or solution from a related project?
- User comparison, feedback, or question?
- Research, article, or talk with relevant insights?
Gather context — URL, date, key details. If the user provides a GitHub URL, fetch the issue/PR for full context.
Classify — determine the right file:
- Existing reference file → append a new entry
- New topic → create .feature-radar/references/{topic}.md
Assess impact:

<HARD-GATE> Before writing to references/, assess impact by answering ALL of these: - New opportunity or feature idea? → state yes/no, suggest file if yes - Way to enhance existing features? → state yes/no, suggest update if yes - Ecosystem trend? → state yes/no, suggest specs/ecosystem-trends.md update if yes </HARD-GATE> 5. **Checkpoint** — State what was written and the impact assessment results. Ask: "I've updated `references/{topic}.md`. Does this look correct, or should I adjust anything?" Wait for user confirmation before proceeding. 6. **Update base.md** — increment the references count in Tracking Summary

File Format

Use the format defined in ../feature-radar/references/SPEC.md § 3.5 (references/{topic}.md).

Naming Convention

Name by the subject being tracked, not the event:

Good: vercel-skills-ecosystem.md, agent-path-conventions.md, cli-ux-patterns.md
Bad: 2026-02-18-update.md, interesting-finding.md

Guidelines

Always cite source URLs and dates for traceability.
Append new entries chronologically to existing files — don't create a new file per observation.
Be objective. Record what happened, then assess implications separately.
If the observation reveals an unmet need or innovation opportunity, proactively suggest creating an opportunity.
Look for creative inspiration, not just feature gaps — how others solve problems can spark new ideas.

Example Output

→ Appended to references/cli-ux-patterns.md
→ Suggested: new opportunity "interactive config wizard"
→ Updated base.md: references 3 → 3 (appended, not new file)

Completion Summary

Follow the template in ../feature-radar/references/DIRECTIVES.md, with skill name "Ref Complete".

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.