forked from bchanot/claude
optimize analyze-system: d1 triggers in skill description + d2 ordered TASKS mapped to OUTPUT sections in analyzer
This commit is contained in:
+6
-7
@@ -25,14 +25,13 @@ Produce a clear analysis without proposing solutions.
|
||||
|
||||
---
|
||||
|
||||
## TASKS
|
||||
## TASKS (in order — each step feeds the OUTPUT section named)
|
||||
|
||||
- Identify relevant parts of the codebase
|
||||
- Understand current behavior
|
||||
- List dependencies
|
||||
- Highlight constraints
|
||||
- Detect risks
|
||||
- Identify ambiguities
|
||||
1. **Locate** — find the relevant parts of the codebase (Glob/Grep from the target) → file list
|
||||
2. **Understand** — read them; describe current behavior as-is → CONTEXT, KEY COMPONENTS
|
||||
3. **Map dependencies** — imports, call sites, data flow in/out → KEY COMPONENTS roles
|
||||
4. **Constrain** — invariants, contracts, conventions the code obeys → CONSTRAINTS
|
||||
5. **Assess** — risks with probability, then ambiguities → RISKS, OPEN QUESTIONS
|
||||
|
||||
---
|
||||
|
||||
|
||||
Reference in New Issue
Block a user