Audit paper drafts for logical consistency, compliance, and academic integrity (Triangulation Matrix).
Scanned 9/12/2026
Install to Claude Code
npx -y skills add aibot88/sec_skill_store --skill paper-audit --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Paper Audit?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/aibot88-paper-audit)More formats (shields.io, HTML) on the badges page.
---
name: paper-audit
description: Audit paper drafts for logical consistency, compliance, and academic integrity (Triangulation Matrix).
allowed-tools:
- Read
- Write
- Glob
---
# Paper Audit Skill
**Goal**: Simulate a ruthless reviewer to check Intent-Implementation consistency and logical triangulation.
## Workflow
1. **Read Input**: Read the user's paper draft (or specific sections).
2. **Load Standards**: Read `paper-audit/CHECKLIST.md`.
3. **Execute Audit Layers**:
* **Layer 1: Compliance Check**:
* Check against `CHECKLIST.md`.
* Output Pass/Fail/Warn table.
* **Layer 2: The Triangulation Matrix (Strict Verification)**:
* Compare **Claims** (Abstract/Intro) vs. **Evidence** (Experiments/Results).
* **Output Table**:
* Column 1: The Promise (Specific claims).
* Column 2: The Proof (Figure/Table #).
* Column 3: The Gap (Exact match or weak proxy?).
* Verdict: ✅ Validated / ⚠️ Weak Support / 🔴 Hallucination.
* **Layer 3: Defensive Audit (Logic Stress Test)**:
* **Baseline Manipulation**: Are weak baselines used?
* **Ablation Amnesia**: Are modules justified by ablations?
* **Figure-Text Consistency**: Do numbers match?
* **Variable Continuity**: Are symbols consistent?
* **Recency Check**: Are >50% citations from last 1-3 years?
* **Layer 4: The Decision**:
* List top 3 Fatal Flaws.
* Mock Decision: Accept / Weak Accept / Major Revision / Reject.
4. **Output**:
* Save full report to `psmfiles/paper_audit_report.md`.
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!