Skip to content
Back to skills

Intent Extraction

ASecurity

Clarify an underspecified outcome or conduct a requirements interview when material choices remain unresolved.

  • 3 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 13, 2026
ai-agents

Security analysis

A100/100

Scanned October 13, 2026

npx -y skills add sujith011/agentic-pro-skills --skill intent-extraction --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Intent Extraction?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Intent Extraction
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/sujith011-intent-extraction/badge)](https://www.skillsdirectory.com/skills/sujith011-intent-extraction)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: intent-extraction
description: "Clarify an underspecified outcome or conduct a requirements interview when material choices remain unresolved."
---

# Intent Extraction
Clarify what changes the implementation, then proceed within the user's authorized scope.

## Process
1. Read the request, relevant repository context, and decisions already made.
2. State the outcome, success criteria, constraints, and material unknowns. Avoid invented confidence percentages.
3. Ask one focused question at a time when different answers would change scope, cost, data handling, or behavior. Attach a reasonable recommendation.
4. Treat explicit delegation such as "use your judgment" as permission to decide routine details. Record the assumption and continue.
5. Restate the resulting scope briefly. Existing authorization is sufficient; request confirmation only for an unresolved consequential choice or a boundary requiring it.

## Example
"Add CSV export" may require asking which records the user can export. Delimiter choice can follow existing conventions without another approval.

## Stop conditions
Do not guess an access rule, destructive action, or materially different product outcome. Continue independent work while that question is pending.

## Evidence
- [ ] Outcome and acceptance criteria are concrete.
- [ ] Material unknowns are resolved or explicitly pending.
- [ ] Routine assumptions and existing authorization are preserved.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.