Skip to content
Back to skills

Human Voice

ASecurity

Extract and profile human typing cadence, style markers, and tone across AI tool logs. Trigger via /human-voice.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 20, 2026
ai-agentspythongobashapi

Works with

  • claude code
  • cursor
  • cli
  • api

Security analysis

A100/100

Pro scans all 5 files and shows the line behind each finding

Scanned September 20, 2026

npx -y skills add ksprashu/agent-skill-forge --skill human-voice --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Human Voice?

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

Security grade badge for Human Voice
[![Security: A β€” Skills Directory](https://www.skillsdirectory.com/api/skills/ksprashu-human-voice/badge)](https://www.skillsdirectory.com/skills/ksprashu-human-voice)

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: human-voice
description: Extract and profile human typing cadence, style markers, and tone across AI tool logs. Trigger via /human-voice.
disable-model-invocation: true
---

# Human-Voice: Persona & Speech Cadence Profiler

Scan developer conversation logs, scrub PII, and extract authentic writing style markers.

---

## 🎯 Goal
Analyze prompt histories across developer tools (Antigravity, Claude Code, Gemini CLI, Cursor) and generate reusable voice profiles without leaking credentials.

---

## πŸ“‹ Step-by-Step Workflow

1. **Run Extraction Pipeline**:
   ```bash
   python3 scripts/extract_voice.py
   ```
2. **Automatic PII Scrubbing**: The script sanitizes private emails, corporate domains, and API tokens.
3. **Inspect Output**: Review generated `output/voice_and_tone.md` and `output/golden_examples.md`.
4. **Deploy Local Profile**: Copy to `references/voice_and_tone.local.md` for local writing personalization.

---

## πŸ’‘ Concrete Example

### Extracted Style Marker Fixture (`voice_and_tone.md`)
```markdown
# Writing Style Profile

## Core Cadence
* Direct, command-driven sentences ("Subtract before you add").
* Prefers bulleted checklists over narrative paragraphs.
* Banned filler phrases: "delve into", "tapestry", "it's worth noting".
* Product linking: Always uses dual links `[Product](landing) ([Docs](docs))`.
```

---

## 🚫 Hard Constraints

*   **NEVER** commit unscrubbed chat logs with real API keys or personal emails.
*   **NEVER** output private corporate intranet URLs without placeholder redaction.

Files in this skill

  • EXAMPLES.md3.2 KB
  • README.md291 B
  • REFERENCE.md4.5 KB
  • SKILL.md1.5 KB
  • scripts/extract_voice.py24.5 KB

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.