multiAI Summary Pending
descriptive-action
Use when the user asks to describe, summarize, analyze, compare, explain, or report on something (text, data, events, systems) without asking for recommendations or next steps.
231 stars
Installation
Claude Code / Cursor / Codex
$curl -o ~/.claude/skills/descriptive-action/SKILL.md --create-dirs "https://raw.githubusercontent.com/aiskillstore/marketplace/main/skills/bellabe/descriptive-action/SKILL.md"
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/descriptive-action/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How descriptive-action Compares
| Feature / Agent | descriptive-action | Standard Approach |
|---|---|---|
| Platform Support | multi | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Use when the user asks to describe, summarize, analyze, compare, explain, or report on something (text, data, events, systems) without asking for recommendations or next steps.
Which AI agents support this skill?
This skill is compatible with multi.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Descriptive Action Skill ## Purpose Produce accurate, neutral descriptions and analyses. Do not prescribe actions unless explicitly requested. ## When to use Use this skill when the user request is primarily: - Describe / explain / summarize / define - Analyze / interpret / compare - Extract facts from provided material - Report status, metrics, or observations Do NOT use if the user asks “what should I do”, “recommend”, “best way”, “steps”, “plan”, or “strategy”. ## Operating rules 1. Stay observational: focus on what is true in the input and what can be inferred safely. 2. Separate facts from interpretation: - Facts: directly supported by the provided input. - Inferences: clearly labeled. 3. If key information is missing, state what’s missing and proceed with bounded analysis. 4. Avoid normative language. 5. Prefer structure over prose. ## Inputs - Text, data, artifacts, or systems to describe - Any stated constraints (scope, timeframe, audience) ## Outputs Structured descriptive analysis using the format below. ### Summary - 3–6 bullets capturing the main points. ### Details - Organized sections (background, findings, trends, constraints). ### Evidence - Brief references to supporting input. ### Open questions - Unknowns limiting confidence.