← Back to Skills Marketplace
Voice-Matched Content System
by
Brian Wagner
· GitHub ↗
· v2.0.0
572
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install voice-matched-content-system
Description
Extract someone's authentic writing voice from samples, build a complete Voice DNA profile, then generate content that sounds like them — not AI. Covers conf...
Usage Guidance
This skill's functionality is coherent with its description, but before installing or using it you should: (1) confirm where the generated Voice DNA will be stored, who can access it, and how to delete it; (2) avoid uploading others' communications unless you have explicit permission; redact PII, financial or medical details from samples; (3) ask the publisher how the optional 5-minute conversation is captured/transcribed and whether audio leaves your device; (4) consider legal/ethical risks of producing content that deliberately mimics another person's voice (impersonation risk); and (5) request safeguards (consent checks, retention policy, audit logging) and sample outputs so you can evaluate accuracy and potential for misuse. If these questions are unanswered, treat the skill cautiously or prefer a manual workflow that keeps data local and auditable.
Capability Analysis
Type: OpenClaw Skill
Name: voice-matched-content-system
Version: 2.0.0
The skill is designed for text analysis and generation, specifically to create a 'Voice DNA Profile' and then generate content matching that voice. All instructions in SKILL.md are directed at the AI agent's internal reasoning, text processing, and content generation capabilities. There are no instructions for executing external commands, accessing sensitive files, making network requests, or any other behavior indicative of data exfiltration, persistence, or unauthorized system interaction. The prompt injection surface is used for legitimate skill-specific instructions, not for subverting the agent for malicious purposes.
Capability Assessment
Purpose & Capability
The name/description (extract voice, build Voice DNA, generate content) align with the SKILL.md instructions. However, the skill promises a "permanent reference" Voice DNA for future use but provides no mechanism for storing or protecting that profile (no install, no storage path, no API). That is a capability/expectation mismatch worth clarifying.
Instruction Scope
Instructions ask the agent to collect personal writing samples (3–10) and optionally a 5-minute conversation, then extract many fine-grained features (e.g., 8–10 transition phrases). There are three issues: (1) the minimum sample count is likely insufficient for some extraction goals (it still requires 8–10 transition phrases); (2) it does not provide guidance on consent, redaction, or handling of sensitive/PII content that users may submit; (3) it is ambiguous how spoken conversation is captured/transcribed (no guidance on audio upload, recording permissions, or required tools). The instructions otherwise remain within the stated purpose and do not request unrelated system access.
Install Mechanism
Instruction-only skill with no install spec and no code files — lowest-risk from an install perspective. Nothing is downloaded or written by an installer.
Credentials
The skill requests no environment variables, credentials, or system paths (proportionate technically). However, it explicitly requires users to supply potentially sensitive personal content (emails, transcripts, texts). That user-provided data can contain PII or copyrighted material; the SKILL.md offers no guidance on redaction, consent if samples include other people's communications, or retention policy.
Persistence & Privilege
The skill is not marked always:true and does not request elevated privileges. That said, it promises a reusable 'Voice DNA' profile for future content but gives no information about where/how that profile is persisted, who can access it, or how to delete it. This gap is a privacy/persistence concern (not a technical privilege request).
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install voice-matched-content-system - After installation, invoke the skill by name or use
/voice-matched-content-system - Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.0.0
v2 — refined Voice DNA extraction with confidence calibration, energy mapping, and platform-specific voice adaptation.
Metadata
Frequently Asked Questions
What is Voice-Matched Content System?
Extract someone's authentic writing voice from samples, build a complete Voice DNA profile, then generate content that sounds like them — not AI. Covers conf... It is an AI Agent Skill for Claude Code / OpenClaw, with 572 downloads so far.
How do I install Voice-Matched Content System?
Run "/install voice-matched-content-system" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Voice-Matched Content System free?
Yes, Voice-Matched Content System is completely free (open-source). You can download, install and use it at no cost.
Which platforms does Voice-Matched Content System support?
Voice-Matched Content System is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Voice-Matched Content System?
It is built and maintained by Brian Wagner (@brianrwagner); the current version is v2.0.0.
More Skills