code-impact-analysis
Analyze code impact before refactors, behavior changes, API changes, and cross-module edits. Use when missing a caller, schema consumer, route, or state dependency could break the product.
Datos de origen
- Repositorio
- 777genius/universal-agent-plugins-registry
- Última actividad en el origen
- 7 de agosto de 2026 a las 13:57
- Idioma detectado de SKILL.md
- inglés
- Estrellas
- 3
- Forks
- 0
Opciones de instalación
De forma predeterminada está seleccionado el prompt que primero revisa el origen. Puedes cambiar a un comando directo o descargar una copia local.
Revisa los archivos de origen
Lee SKILL.md y los archivos complementarios que muestra SkillsMP antes de decidir si quieres instalarlo.
Mostrando SKILL.md
SKILL.md
Instrucciones de origen · Vista previa de solo lectura- name
- code-impact-analysis
- description
- Analyze code impact before refactors, behavior changes, API changes, and cross-module edits. Use when missing a caller, schema consumer, route, or state dependency could break the product.
- license
- Apache-2.0
- compatibility
- Requires a local code workspace. Uses rg first; Semble, Serena, and CodeGraphContext are optional.
# Code Impact Analysis
Build a verified impact set before editing shared behavior or public contracts.
1. Locate the definition with an LSP tool or `rg`.
2. Find exact call sites and serialized field names with `rg`.
3. Find type-aware references with Serena or the language server.
4. Ask CGC for callers, callees, dependencies, and complexity when indexed.
5. Use Semble to find conceptually similar patterns that exact search may miss.
6. Read every high-risk caller and decide the minimum edit set.
7. Identify focused tests, type checks, migrations, and user-visible verification.
Examples:
```bash
rg -n "symbolName\\(" .
uvx --from semble==0.5.4 semble search "similar usage of symbolName" .
cgc analyze callers symbolName --context "$(basename "$PWD")"
cgc analyze calls symbolName --context "$(basename "$PWD")"
```
When tools disagree, inspect the conflicting results. Generated code, dynamic
dispatch, framework conventions, or an incomplete index can explain the gap.
Never reduce the impact set solely because a graph or semantic search returned
few results.
Ver en GitHub