Smithery is now a part of Arcade.dev! Read more in our announcement here
Routes tasks to reasoning frameworks and command patterns in commands-db...
Routes tasks to appropriate reasoning frameworks and command patterns.
Activate when task involves:
Primary action verbs for development tasks.
| Command | Trigger Keywords | Purpose |
|---|---|---|
sc:analyze |
analyze, evaluate, assess, audit | Deep analysis |
sc:build |
build, compile, package | Build operations |
sc:cleanup |
clean, organize, declutter | Code cleanup |
sc:design |
design, architect, plan | Architecture |
sc:document |
document, readme, guide | Documentation |
sc:estimate |
estimate, cost, time | Estimation |
sc:explain |
explain, clarify, describe | Explanations |
sc:git |
commit, push, merge | Git operations |
sc:implement |
implement, code, feature | Implementation |
sc:improve |
improve, enhance, optimize | Improvement |
sc:load |
load, context, import | Context loading |
sc:spawn |
spawn, delegate, parallel | Delegation |
sc:task |
task, complex, persistent | Complex tasks |
sc:test |
test, verify, validate | Testing |
sc:troubleshoot |
debug, fix, diagnose | Debugging |
sc:workflow |
workflow, prd, process | Workflows |
| Module | Purpose |
|---|---|
bmad:brainstorm |
Structured brainstorming |
bmad:party-mode |
Multi-agent ideation |
bmad:prd |
Product requirements |
bmad:architecture |
System architecture |
# Use reasoning-index for semantic routing
reasoning-index search "{user_intent}"
reasoning-index suggest "{task_description}"
# Example output:
# {
# "commands": ["sc:implement", "sc:design"],
# "skills": ["test-driven-development"],
# "reasoning_chain": ["sc:design", "sc:implement", "sc:test"]
# }
Command Task Detected
│
├── Implementation?
│ ├── New feature? → sc:implement
│ ├── Build? → sc:build
│ └── Improvement? → sc:improve
│
├── Analysis?
│ ├── Code review? → sc:analyze
│ ├── Debugging? → sc:troubleshoot
│ └── Testing? → sc:test
│
├── Design?
│ ├── Architecture? → sc:design
│ ├── Brainstorm? → bmad:brainstorm
│ └── PRD? → bmad:prd
│
├── Documentation?
│ ├── README? → sc:document
│ ├── Explain? → sc:explain
│ └── Guide? → sc:document
│
└── Operations?
├── Git? → sc:git
├── Cleanup? → sc:cleanup
└── Complex? → sc:task
# Search commands
reasoning-index search "debug authentication"
# Get reasoning chain suggestion
reasoning-index suggest "implement user login"
# List all SC commands
reasoning-index list sc
# Build knowledge graph
reasoning-index graph "optimize performance" --depth 2
Design → Implement → Test → Document
sc:design → sc:implement → sc:test → sc:document
