# Doncheli Voice

> Convert voice dictation into structured specs, review comments, or brainstorming notes. Activate when user mentions "voice", "dictate", "speak", "voice mode", "I'll dictate", "voice input".

- **Type:** Skill
- **Install:** `agentstack add skill-doncheli-don-cheli-sdd-doncheli-voice`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [doncheli](https://agentstack.voostack.com/s/doncheli)
- **Installs:** 0
- **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml)
- **Latest version:** 0.1.0
- **License:** Apache-2.0
- **Upstream author:** [doncheli](https://github.com/doncheli)
- **Source:** https://github.com/doncheli/don-cheli-sdd/tree/main/.agent/skills/doncheli-voice
- **Website:** https://doncheli.tv

## Install

```sh
agentstack add skill-doncheli-don-cheli-sdd-doncheli-voice
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Don Cheli: Voice Mode Integration

## Instructions

1. Prompt the user to activate Claude Code Voice Mode if not already active
2. Accept transcribed voice input as raw text
3. Detect the intended mode from context or user flag:
   - `spec` — convert to Gherkin Feature/Scenario format
   - `review` — convert to numbered review comments with severity
   - `brainstorm` — organize ideas into grouped bullet points
4. Clean the transcription: remove filler words, fix punctuation, resolve run-on sentences
5. Structure the output according to the detected mode
6. Mark genuinely ambiguous segments with `[unclear: original text]` — do not invent content
7. Present the structured output and ask for confirmation before saving
8. On confirmation, save to the appropriate location:
   - spec → `.dc/specs/.feature`
   - review → `.dc/reviews/-review.md`
   - brainstorm → `.dc/notas/-brainstorm.md`
9. Suggest relevant follow-up commands: `/dc:validar-spec`, `/dc:debate`, `/dc:clarificar`

## Quality Gate

- Never save without user confirmation
- If the transcription is shorter than 3 meaningful sentences, ask if the user wants to continue dictating
- Preserve technical terms exactly as dictated — do not paraphrase class names, endpoints, or identifiers

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [doncheli](https://github.com/doncheli)
- **Source:** [doncheli/don-cheli-sdd](https://github.com/doncheli/don-cheli-sdd)
- **License:** Apache-2.0
- **Homepage:** https://doncheli.tv

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-doncheli-don-cheli-sdd-doncheli-voice
- Seller: https://agentstack.voostack.com/s/doncheli
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.