# Domain Context

> {{PLUGIN_NAME}} Context Skill

- Skill: `brite-nites/domain-context` (Agent Skill)
- Install (CLI): `npx skillmds@latest add brite-nites/domain-context`
- Raw SKILL.md: https://api.skillmd.com/api/skills/brite-nites/domain-context/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: Brite-Nites (https://skillmd.com/u/brite-nites)
- Updated: 2026-09-21
- Page: https://skillmd.com/skills/brite-nites/domain-context

---


# {{PLUGIN_NAME}} Context Skill

This is the foundational context-skill for the {{PLUGIN_NAME}} domain plugin. It creates and maintains the project-specific context document that all other skills in this plugin read before acting.

## When This Activates

- **Auto-triggered** by `project-start` when the `{{TRAIT}}` trait is detected and this plugin is installed
- **Manually invocable** via `/{{PLUGIN_SLUG}}:domain-context` to refresh the context doc

## What It Produces

**Output file:** `docs/{{DOMAIN}}-context.md`

The context doc contains foundational {{PLUGIN_NAME}} knowledge specific to this project:
- Project-specific domain context gathered during the project-start interview
- SoR-enriched data (if MCP tools are available)
- Frontmatter with `last_refreshed` and `refresh_cadence` for staleness tracking

### Context Doc Frontmatter

```yaml
---
domain: {{DOMAIN}}
trait: {{TRAIT}}
last_refreshed: YYYY-MM-DD
refresh_cadence: quarterly
generated_by: {{SKILL_NAME}}
---
```

**CRITICAL:** Use `last_refreshed`, not `last_generated` — session-start parses this exact key.

## SoR Integration

Context-skills query two tiers of Systems of Record to enrich the context doc:

### Tier 1 — Handbook (universal, currently unavailable)

The Brite Handbook (company-wide knowledge: brand guidelines, ICP definitions, coding standards, competitive positioning, org structure) was previously queryable as a universal SoR via Context7. **Context7 was removed (BC-11891); ADR-001 is Withdrawn.** No replacement is wired yet — a future gbrain-based integration is anticipated. For now, skip Tier 1 and proceed to Tier 2.

When the handbook query path is restored, this section will document the new mechanism (likely `resolve` → `query` style against gbrain).

### Tier 2 — Domain MCP (query first today; query second when Tier 1 returns)

If a domain-specific MCP tool is available (see spec for trait-to-SoR mapping), query it for live data:

1. Check domain MCP availability
2. Query domain SoR for domain-relevant data
3. **MANDATORY: Follow Data Safety rules** before writing any SoR data — see `docs/designs/BC-1966-context-skill-standard.md` § Data Safety for: newline stripping, character allowlist, field/list caps, frontmatter exclusion, blockquote wrapping
4. Write enriched sections to context doc
5. Record query metadata in `## SoR Sources`

### Fallback behavior

- **Both available:** Full enrichment — handbook + domain SoR + interview data
- **Handbook only:** Handbook-enriched context, `<!-- needs-enrichment -->` on domain SoR sections
- **Domain SoR only:** SoR-enriched but missing handbook company context
- **Neither available:** Interview data only

## How Sibling Skills Use This

All other skills in this plugin MUST read `docs/{{DOMAIN}}-context.md` at the start of their execution. If the file doesn't exist, warn and proceed with reduced context — never hard-fail.

## Specification

See `docs/designs/BC-1966-context-skill-standard.md` for the full context-skill standard, including:
- Required frontmatter schema
- Content format and budget (~80-200 lines)
- SoR query pattern and fallback tiers
- Cross-plugin reference table