# Build Your LinkedIn Writing Voice

Generic outreach drafts often lose the sender's writing style. This skill builds a voice card from real messages and checks new drafts against approved examples.

## Install

```bash
npx skillstore add bereach-ai/linkedin-voice
```

## Metadata

- Status: approved
- Slug: bereach-ai-linkedin-voice
- Skillstore revision: r1
- Version status: missing
- Tree hash: e454752e44dfab82989c9dbf39d03f784fb29a7ec8a427895ac74b7a7b1bb612
- Author: bereach-ai
- GitHub username: bereach-ai
- Repository: https://github.com/bereach-ai/linkedin-prospecting-skills/tree/6967fde3a1b93fd84e2dfa6be39f56113d79ceca/skills/linkedin-voice
- Ref: 0b5375d5ff31ef8f8f23150e2d3ac5978ea0441f
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: confirmation\_required
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/bereach-ai-linkedin-voice
- Manifest: https://skillstore.pages.dev/api/skills/bereach-ai-linkedin-voice/manifest

## Capabilities

- Extract greetings, sign-offs, punctuation, pronoun habits, sentence lengths, and signature phrases from three to five messages.
- Record quoted evidence for voice-card fields and keep unsupported fields empty.
- Separate consistent writing habits from tone differences across invitations, first messages, follow-ups, and replies.
- Apply companion message frameworks to revise examples while retaining the sender's identity and stated writing restrictions.
- Save approved examples and writing guidelines as a language-specific BeReach playbook when the required tools are available.
- Check drafts for voice consistency, forbidden phrases, address form, and invented facts or offers.

## Use Cases

- Keep founder outreach personal: Turn a founder's own successful conversations into a reusable writing profile for new prospect drafts.
- Standardize individual seller profiles: Create separate evidence-based voice cards so representatives retain their own greetings, vocabulary, and message tone.
- Maintain language-specific client voices: Build approved cards for each client's language and review drafts against their address forms and forbidden wording.

## Prompt Templates

### Analyze my writing samples

```
Build a draft voice card from these three messages I wrote: [paste messages]. Quote evidence for each field and leave unsupported fields empty.
```

### Check a draft against my voice

```
Compare this draft with my approved voice card: [paste both]. Fix mismatches in greeting, address form, punctuation, and forbidden wording.
```

### Create a reference and test message

```
Use my samples, a real recipient, and verified facts: [paste inputs]. Apply the available companion frameworks and explain revisions. Draft only; do not send.
```

### Update an approved saved playbook

```
Review my saved playbook and these repeated corrections: [paste corrections]. Propose scoped changes and request approval before saving. Ask before any chargeable tool use.
```

## Limitations

- Full operation depends on BeReach tools and companion skills that are not included in these four files.
- A small sample supports style guidance, not reliable authorship identification or guaranteed outreach results.
- The files define drafting and review guidance, not a standalone application or message-sending integration.
- The current instructions suppress cost discussion; resolve this audit concern before relying on potentially chargeable briefings.

## Best Practices

- Provide varied messages you wrote and identify which received substantive replies.
- Review quoted evidence and approve revised samples before saving a playbook containing conversation text.
- Confirm tool costs, data access, and save scope before approving BeReach operations.

## Anti Patterns

- Using unedited generated messages as proof of your personal writing style.
- Treating a few positive replies as proof that a message framework improves response rates.
- Approving tool usage without understanding costs or which conversation excerpts will be stored.

## Security Audit

- Audited at: 2026-11-14T17:43:56.189\+00:00
- Summary: All 36 static findings are false positives involving Markdown formatting or ordinary prose. No evidence found of shell execution, weak cryptography, or system reconnaissance. A separate instruction suppresses user-requested cost information while requiring full briefing, creating a budget-control and transparency risk.

## Stats

- Views: 0
- Downloads: 3
- Favorites: 0
- Popularity score: 0
