# Brand Context

> Brand Context Sync

- Skill: `matteotitta/brand-context` (Agent Skill)
- Install (CLI): `npx skillmds@latest add matteotitta/brand-context`
- Raw SKILL.md: https://api.skillmd.com/api/skills/matteotitta/brand-context/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: matteotitta (https://skillmd.com/u/matteotitta)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/matteotitta/brand-context

---


# Brand Context Sync

Synthesise the "Voice & Messaging" section for a client's CLAUDE.md from existing upstream skill outputs. This is a meta-skill — it reads files already in the project folder, it does not perform new research.

The purpose: every content skill auto-loads the client CLAUDE.md. By adding a standardised "Voice & Messaging" section, content skills get instant access to voice rules, messaging anchors, and competitive context without re-asking the user.

For full 3-phase process + DESIGN.md awareness rules → the premium reference.

---

## Claude Code Triggers

**Invoke this skill when:**
- "sync brand context for [client]"
- "update voice section for [client]"
- "refresh client context"
- "add voice & messaging to CLAUDE.md"
- After completing `/tov-guidelines` for any client

**Do NOT invoke when:**
- User wants to run TOV analysis from scratch → use `/tov-guidelines`
- User wants full competitor research → use `/competitor-research`
- User wants to update project scope or file references in CLAUDE.md → edit manually

---

## Input Requirements

### Required
| Input | Description | Source |
|-------|-------------|--------|
| **Client project folder** | Must be inside `projects/consulting/{slug}/` | User context or explicit path |
| **Client CLAUDE.md** | The file to update | `projects/consulting/{slug}/CLAUDE.md` |

### Optional (improve quality)
| Input | How it helps |
|-------|--------------|
| TOV guidelines output | Voice rules (Use/Avoid lists) |
| Brand guidelines output | Visual identity, vocabulary |
| Competitor research aggregate | Competitive quick-ref |
| ICP research output | ICP description, key people |
| Product messaging output | Messaging anchors (value prop, differentiators, CTA) |
| Expert POV output | Author voice and positioning |

### Validation checklist
Before proceeding, verify:
- [ ] Client CLAUDE.md exists and is readable
- [ ] At least one upstream file exists (brand/, competitors/, icp/, messaging/)
- [ ] Template file exists at `VOICE-MESSAGING-TEMPLATE.md`

If no upstream files exist: tell the user "No brand files found. Run `/tov-guidelines` first to generate source material."

---

## Process at a glance

| Phase | Purpose | Output |
|-------|---------|--------|
| **1. Locate** | Scan project folder, determine CREATE vs UPDATE | File inventory + mode |
| **2. Extract** | Pull voice rules, vocabulary, messaging anchors, competitive quick-ref, key people | Section content (or MISSING markers) |
| **3. Write** | Format via template, insert/replace in client CLAUDE.md, verify | Updated CLAUDE.md |

For full per-phase steps, MISSING-marker conventions, and DESIGN.md awareness → the premium reference.

---

## Anti-Hallucination Guardrails

1. **Verbatim voice rules.** Copy Use/Avoid rules exactly from TOV. Don't paraphrase, generalise, or add rules not in source.
2. **No invented messaging.** If positioning/messaging output doesn't exist, mark as MISSING. Don't invent value props or differentiators.
3. **No invented competitors.** Only include competitors from competitor-research files. Don't add from general knowledge.
4. **Source tracing.** Every section must reference its source file.
5. **Mark unknowns explicitly.** Use `[MISSING — run /[skill-name] first]` for any section that can't be populated.

---

## Quality Checklist (pre-delivery)

### Content quality
- [ ] Voice rules are verbatim from source (not paraphrased)
- [ ] Max 12 Use + 12 Avoid rules (not more)
- [ ] Messaging anchors traced to positioning/messaging output
- [ ] Competitive quick-ref matches aggregate analysis
- [ ] MISSING sections clearly marked with the skill to run

### Format quality
- [ ] Section header is exactly `## Voice & Messaging (auto-loaded by content skills)`
- [ ] Subsection headers match template (Identity, Voice rules, Vocabulary, etc.)
- [ ] Arrows (→) used for differentiator lists
- [ ] Source file paths are relative to project root

### Completeness
- [ ] All 7 subsections present (even if some are MISSING)
- [ ] Sources subsection lists all files that were read
- [ ] No other CLAUDE.md sections modified

---

## Integration with Other Skills

| Skill | Relationship | Usage |
|-------|--------------|-------|
| **tov-guidelines** | Primary upstream | Voice rules + vocabulary |
| **brand-kit** | Upstream | Visual identity context |
| **competitor-research** | Upstream | Competitive quick-ref |
| **icp-behavioural** | Upstream | ICP description |
| **product-messaging** | Upstream | Messaging anchors |
| **expert-pov** | Upstream | Author voice |
| **All content skills** | Downstream consumers | Auto-load the Voice & Messaging section from client CLAUDE.md |

---