Use when stress-testing documentation for ambiguities, contradictions, and developer confusion. Triggers on 'elicit', 'stress-test', 'challenge the spec', 'find gaps', 'interrogate docs', 'what did we miss', or 'semantic review'.
From pmnpx claudepluginhub etusdigital/etus-plugins --plugin pmThis skill uses the workspace's default tool permissions.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.
Guides slash command development for Claude Code: structure, YAML frontmatter, dynamic arguments, bash execution, user interactions, organization, and best practices.
Interrogate existing documentation to surface ambiguities, contradictions, missing boundaries, and hidden assumptions BEFORE they reach implementation. This skill does NOT create new product artifacts — it produces a findings report only.
[feature-slug] provided → scope to docs/ets/projects/{project-slug}/features/{feature-slug}/docs/ets/projects/{project-slug}/--focus provided → run only the matching technique subsetFor each feature or user story, adopt the perspective of a developer who has never seen the codebase:
Compare overlapping definitions across documents:
NOT NULL in DDL but optional in API request schema → EL-C-#max_length: 255 in data-dictionary but no validation in API → EL-C-#Search all documents for subjective language that should be quantified:
Read all NG-# (non-goal) declarations from product-vision, feature-brief, or PRD:
Collect all EDGE-# IDs from opportunity-pack, feature-spec, and user-stories:
For each happy-path scenario (UC-#, US-# with happy-path Given/When/Then):
Scan all documents for permission-related claims:
For each entity that has lifecycle states (e.g., order: draft > pending > confirmed > shipped > delivered):
| Flag | Techniques |
|---|---|
--focus=permissions | #7 (deep), #1 (permission questions only) |
--focus=failures | #5, #6, #8, #1 (failure questions only) |
--focus=consistency | #2, #4, #3 |
| (no flag) | All 8 techniques |
The report is saved to docs/ets/projects/{project-slug}/state/reports/elicit-report.md.
# Elicit Report — {project-slug} [{feature-slug | "full project"}]
Generated: {timestamp}
Scope: {path}
Focus: {all | permissions | failures | consistency}
## Summary
| Category | Count | Critical |
|----------|-------|----------|
| EL-A (Ambiguities) | N | N |
| EL-C (Contradictions) | N | N |
| EL-B (Missing Boundaries) | N | N |
| EL-H (Hidden Assumptions) | N | N |
| EL-E (Unresolved Edges) | N | N |
| EL-D (Developer Questions) | N | N |
| **Total** | **N** | **N** |
## Findings
### EL-A-001: {title}
- **Location:** {document} > {section}
- **Issue:** {description}
- **Severity:** HIGH | MEDIUM | LOW
- **Suggested Resolution:** {actionable fix}
- **Evidence:** "{quote from document}"
### EL-C-001: {title}
...
## Recommendations
1. {Highest priority fix}
2. {Second priority fix}
...
## Technique Coverage
| Technique | Ran | Findings | Notes |
|-----------|-----|----------|-------|
| 1. Developer Simulation | Yes/No | N | |
| 2. Cross-Doc Consistency | Yes/No | N | |
| 3. Vague Quantifier Scan | Yes/No | N | |
| 4. NG-# Violation Scan | Yes/No | N | |
| 5. EDGE-# Resolution Audit | Yes/No | N | |
| 6. Missing Error Siblings | Yes/No | N | |
| 7. Permission Matrix Gap | Yes/No | N | |
| 8. State Machine Completeness | Yes/No | N | |
| Severity | Meaning | Action |
|---|---|---|
| HIGH | Would block or confuse implementation | Must fix before Implementation Readiness Gate |
| MEDIUM | Risk of rework or misinterpretation | Should fix, track as action item |
| LOW | Improvement opportunity | Nice to fix, no urgency |
/elicit to diagnose the gap./elicit to verify consistency survived the change.docs/ets/projects/{project-slug}/state/reports/elicit-report.md