Skip to content
Back to skills

Style Review

ASecurity

Audit existing prose against the user's personal vestige catalog (from their generated output-filter skill). Use on a draft before it ships, to list the vestige hits in it. Returns a hit list with location, pattern, severity, and suggested rewrite. Read-only: report, no modification. Activates on "/style-review", "check for vestiges", "audit this for hedging".

  • 2 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added October 12, 2026
ai-agentsgobashgit

Security analysis

A100/100

Scanned October 12, 2026

npx -y skills add uncovertechtalent/cto-powersuit --skill style-review --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Style Review?

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

Security grade badge for Style Review
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/uncovertechtalent-style-review/badge)](https://www.skillsdirectory.com/skills/uncovertechtalent-style-review)

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: style-review
description: >
  Audit existing prose against the user's personal vestige catalog (from their
  generated output-filter skill). Use on a draft before it ships, to list the
  vestige hits in it. Returns a hit list with location, pattern, severity, and
  suggested rewrite. Read-only: report, no modification. Activates on
  "/style-review", "check for vestiges", "audit this for hedging".
user-invocable: true
---

# Style review

Post-hoc audit against the user's catalog. Companion to `output-filter`, which enforces the catalog at composition time; this skill enforces it on existing text.

## Source of truth

Read the catalog from `~/.claude/skills/output-filter/SKILL.md` before auditing. If no generated catalog exists yet, say so and offer `/calibrate`; auditing against the starter menu alone is allowed but label the report "starter catalog, uncalibrated".

## Scope: whose text is this?

The catalog governs model-to-user output. Before auditing, classify:

- **Model output** (replies, drafts the model wrote): full catalog.
- **The user's own writing** (their posts, emails, articles): do NOT run the catalog unless they explicitly ask for it on their own text. Many writers deliberately use the default register for external audiences; a filter tuned for inbound reading is the wrong tool for outbound writing. When they do ask, say once that the calibration direction differs, then proceed.

Unclear direction: ask.

## Usage

- `/style-review <file path>`: audit the file.
- `/style-review` + pasted prose: audit the paste.
- `/style-review --rewrite <input>`: hit list plus a rewritten version with critical and high hits stripped.

## Output format

```
## Style review: <source>

Total hits: N (critical: N, high: N, medium: N, low: N)

1. Line 12: **<pattern name>**
   > "<quoted specimen>"
   Suggested: <the fix from the catalog entry>

### Summary
- Strongest patterns: <top 2 by count>
- Recommended action: <one line>
```

Severity comes from the user's catalog ordering (entries they confirmed as BLOCK-tier regex rules are critical; prose-only stance rules are medium unless their entry says otherwise). List all hits first, then compute totals from the finished list; never write counts before the hits exist.

## What this is NOT

- Not a grammar or fact checker.
- Not a universal AI-detector. The catalog is fitted to one reader; other readers' legitimate registers will false-positive against it.
- Not a rewrite service by default. `--rewrite` is opt-in; the default report leaves the text alone.

## Overlay

This skill works on the generic defaults in this file. Before you start, check for a private overlay:

```bash
[ -n "$POWERSUIT_OVERLAY" ] && ls -1 "$POWERSUIT_OVERLAY/vestige" 2>/dev/null || true
```

If `$POWERSUIT_OVERLAY/vestige/` exists, read its `README.md` first, then the files in `vestige/style-review/` if that folder exists. Read other overlay files only when that README or this skill names them. Where the overlay and this file disagree, follow the overlay. If the variable is unset or the folder is missing, use the defaults and do not ask for an overlay. If the permission settings block the command, use the defaults and say in your output that the overlay was not checked. Never copy overlay content into this plugin, into its templates or into any public repository.

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.