Extract explicit and latent user goals, resolve ambiguous requests, detect unstated business constraints, and map natural language intents to concrete engineering execution paths.
Scanned 9/6/2026
Install to Claude Code
npx -y skills add saitarrun/Devforge-ai --skill intent-clarification --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Intent Clarification?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/saitarrun-intent-clarification)More formats (shields.io, HTML) on the badges page.
---
name: intent-clarification
description: Extract explicit and latent user goals, resolve ambiguous requests, detect unstated business constraints, and map natural language intents to concrete engineering execution paths.
version: 1.0.0
---
# User Intent Comprehension & Goal Alignment Skill
This skill guides agents to accurately decipher what the user **actually needs** versus what they literally asked for, preventing misalignment before code is written.
---
## 1. Intent Deconstruction Framework (The 4-Layer Lens)
Whenever a user provides a prompt or feature request, analyze it through four concentric layers:
```
[Layer 1: Literal Request] → "Add a login screen"
↓
[Layer 2: Underlying Objective] → User needs secure session management and user identification
↓
[Layer 3: Latent Requirements] → Password reset, OAuth/SSO, session invalidation, CSRF protection
↓
[Layer 4: Non-Functional Drivers]→ Enterprise compliance, low latency, zero-downtime migration
```
---
## 2. Ambiguity Detection & Resolution Matrix
If a prompt contains ambiguous or high-entropy phrasing, immediately classify and address the gap:
| Ambiguity Pattern | Example User Phrasing | Resolution Action |
|---|---|---|
| **Vague Scope** | *"Make it faster"*, *"Make it look modern"* | Quantify targets: Ask for specific p95 latency targets (e.g. <100ms) or UI design references. |
| **Architectural Ambiguity** | *"Connect to external data"* | Determine sync vs async (Webhook vs Polling vs Kafka) and error retry semantics. |
| **Missing Failure Modes** | *"Send email when user signs up"* | Clarify transactional reliability: Should signup fail if the email provider is down? |
| **Implicit Authorization** | *"Admins can manage team members"* | Ask if roles are static or configurable (RBAC vs ABAC). |
---
## 3. High-Fidelity Confirmation Protocol
When clarifying intent with the user:
1. **Replay Understandings with Options**: Present the extracted mental model back in plain language with concrete trade-offs.
2. **Proactively Suggest Industry Defaults**: Never present open-ended confusion; provide a solid (Recommended) option.
3. **Check for Unintended Side-Effects**: State potential impacts on existing workflows (e.g., *"Note: Enabling MFA will require updating the mobile login flow"*).
```markdown
**Intent Confirmation Example:**
"I understand you want to add **team workspace sharing**.
- **Primary Goal**: Allow users to invite team members with specific roles (Admin/Member/Viewer).
- **Recommended Default**: We'll use RBAC with invite links expiring in 48 hours and tenant-isolated data queries.
Does this match your intended vision?"
```
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!