Use Gemini CLI's 1M token context to understand entire codebases in one pass. Full architecture mapping, pattern discovery, and onboarding documentation.
/plugin marketplace add DNYoussef/context-cascade/plugin install dnyoussef-context-cascade@DNYoussef/context-cascadeThis skill inherits all available tools. When active, it can use any tool Claude has access to.
ANTI-PATTERNS.mdREADME.mdexamples/example-1-new-project.mdexamples/example-2-security-audit.mdBefore writing ANY code, you MUST check:
.claude/library/catalog.json.claude/docs/inventories/LIBRARY-PATTERNS-GUIDE.mdD:\Projects\*| Match | Action |
|---|---|
| Library >90% | REUSE directly |
| Library 70-90% | ADAPT minimally |
| Pattern exists | FOLLOW pattern |
| In project | EXTRACT |
| No match | BUILD (add to library after) |
Kaynak dogrulama modu etkin.
Use Gemini CLI's massive 1 million token context window to analyze entire codebases in a single pass. Generate architecture maps, pattern documentation, and onboarding guides.
# Load entire codebase
./scripts/multi-model/gemini-yolo.sh "Analyze architecture" task-id megacontext
# Via delegate.sh
./scripts/multi-model/delegate.sh gemini "Map full architecture" --all-files
# Direct Gemini
bash -lc "gemini --all-files 'Analyze entire codebase and document architecture'"
Common analysis patterns:
| Metric | Value |
|---|---|
| Capacity | 1 million tokens |
| Equivalent | ~1,500 pages |
| Lines of Code | ~30,000 LOC |
| Best for | Projects under 30K LOC |
User: "I just joined this team, help me understand the codebase"
Gemini Analysis:
bash -lc "gemini --all-files 'Create comprehensive onboarding guide including:
1. High-level architecture
2. Key components and their responsibilities
3. Data flow between components
4. Configuration and environment setup
5. Common patterns used'"
Output:
- Architecture: Microservices with API gateway
- Key Services: auth, users, orders, notifications
- Patterns: Repository pattern, Event-driven
- Setup: Docker Compose with 4 services
User: "What would break if we rename User to Account?"
Gemini Analysis:
bash -lc "gemini --all-files 'Identify all files affected by renaming User to Account'"
Output:
- 47 files with direct User references
- 12 database migrations needed
- 8 API endpoints returning User data
- 15 frontend components
- 3 background jobs
User: "Find all places handling sensitive data"
Gemini Analysis:
bash -lc "gemini --all-files 'Security audit: Find all PII handling, auth patterns, and potential vulnerabilities'"
Output:
- PII fields: 12 database columns
- Auth: JWT with refresh tokens
- Concerns: 3 endpoints missing auth middleware
- Logging: 2 instances of sensitive data in logs
gemini --all-files "Document the full system architecture with component interactions"
gemini --all-files "Create a dependency graph showing how all modules relate"
gemini --all-files "Identify all design patterns used and assess consistency"
gemini --all-files "Document all API endpoints with their request/response formats"
META-LOOP PROPOSE PHASE:
|
+---> gemini-codebase-onboard
| |
| +---> Gemini: Load entire codebase (--all-files)
| +---> Gemini: Analyze architecture
| +---> Gemini: Identify impact areas
|
+---> Claude: Synthesize into plan
|
+---> Continue to IMPLEMENT phase
Results stored at:
multi-model/gemini/onboard/{project}/{task_id}Based on real developer feedback:
multi-model-discovery: Find existing solutionsgemini-research: Current info researchcodex-iterative-fix: After understanding, for implementationdocumentation: Generate docs from analysisThis skill should be used when the user asks to "create a hookify rule", "write a hook rule", "configure hookify", "add a hookify rule", or needs guidance on hookify rule syntax and patterns.
Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.