name: ctx description: Create a feature-specific context document and save it under the feature contexts directory for later planning, review, or implementation work. applyTo: []
Skill: /ctx
Purpose: Create context documents for specific features. Feature contexts are stored in the feature directory and linked to the feature document.
Input Formats
Choose any of the following:
/ctx {topic} # Auto-detect FEATURE_ID from chat, provide topic
/ctx feat-2026-01-123 {topic} # FEATURE_ID + context topic
/ctx {topic} --name context-name # Custom output filename
/ctx {topic} --format markdown # Specify output format
When {topic} only: auto-detect FEATURE_ID from chat, create context for topic.
When --name provided: output file is named {context-name}.context.md (otherwise derived from topic).
When --format provided: use specific format (default: markdown with /context-feature structure).
Instructions for user
- Minimal case:
/ctx {topic}— skill infers FEATURE_ID from chat context. - With FEATURE_ID:
/ctx feat-2026-01-123 technical-analysis(explicit feature + topic). - Custom output name:
/ctx my-topic --name my-custom-name(control output filename). - Examples:
/ctx technical-analysis,/ctx architecture-decisions,/ctx domain-knowledge. - ❌
feature-name/ - ❌
custom-feature/
Context Files: {descriptive-name}.context.md or {type}-{topic}.context.md
- ✅
technical-analysis.context.md - ✅
architecture-decisions.context.md - ✅
domain-knowledge.context.md - ❌
context-2025-11-18-topic.md(date not needed in feature contexts) - ❌
topic.md(missing.contextsuffix)
Context to Feature Relationship
One-to-Many: Feature → Contexts
- One feature can have multiple contexts (different topics)
- Each context belongs to exactly one feature
- Contexts are created as needed during implementation or research
- Contexts contain analysis and knowledge not in main feature document
Metadata (optional fields — see artifact-metadata.md ### Context):
kind—researchorbrainstormcategory— free string domain classifier (e.g.technical,architecture,domain-knowledge)tags,status,updated
Cross-document links and trailing refs: see cross-document-links.md.
Research-type context
When the requested context is research (solution options, best practices, 3rd party solutions, broad/deep exploration):
- Loop/auto mode only: Delegate to the Researcher subagent via researcher.md with feature + topic; Researcher produces the context document in
feat-XXX/contexts/{topic}.context.mdusing this command's structure and linking. - If Researcher cannot be launched: Do not produce the context yourself. Tell the user: "This context needs research (options, best practices, 3rd party). Run /research with this feature and topic so the Researcher agent can produce the context document."
Instructions for user
- Provide FEATURE_ID or path to feature doc (e.g.
feat-2025-10-feature-name) - Provide context topic (e.g. "Technical Analysis", "Architecture Decisions")
- Context will be created in
feat-XXX/contexts/{topic}.context.md
Instructions for model
You are the Architect (architect.md). Create context documents for features.
Get current date via date +%Y-%m-%d. Validate feature and context scope. Create context file in feat-XXX/contexts/{topic}.context.md with bidirectional references.
Inputs integration:
- Cite dependencies as markdown links in the context body. Never hand-write or edit trailing
## References/## Used by. - The context file auto-syncs on save. If
syncorcheckreturns an error, show that raw failure to the user.
See also: /research skill, /feat skill
Generateur de Documentation API
Documentation
Genere automatiquement de la documentation API OpenAPI/Swagger.
Rédacteur Technique
Documentation
Rédige de la documentation technique claire selon les meilleurs style guides.
Vérification des revendications documentaires
Documentation
Évaluation en lecture seule des affirmations documentaires soutenues par des preuves. Audite les README, notes de version et guides sans exécuter de commandes ni apporter de modifications.