AgentSkillsCN

file-analysis

为下游的评审与重构工作流提供结构化的上下文支持。适用于在架构评审前了解文件组织方式、探索陌生代码库以梳理结构、估算重构或迁移的范围时使用。切勿在进行一般代码探索时使用此技能——请改用Explore代理。切勿在需要寻找特定模式时使用此技能——请直接使用Grep工具。

SKILL.md
--- frontmatter
name: file-analysis
description: 'Provides structural context for downstream review and refactoring workflows.
  Use when before architecture reviews to understand file organization, exploring
  unfamiliar codebases to map structure, estimating scope for refactoring or migration.
  Do not use when general code exploration - use the Explore agent. DO NOT use when:
  searching for specific patterns - use Grep directly.'
category: workspace-ops
tags:
- files
- structure
- analysis
- codebase
- exploration
tools:
- Bash
- Glob
- Grep
- TodoWrite
complexity: medium
estimated_tokens: 800
dependencies:
- sanctum:shared
- imbue:evidence-logging
version: 1.4.0

File Analysis

When To Use

  • •Before architecture reviews to understand module boundaries and file organization.
  • •When exploring unfamiliar codebases to map structure before making changes.
  • •As input to scope estimation for refactoring or migration work.

When NOT To Use

  • •General code exploration - use the Explore agent
  • •Searching for specific patterns - use Grep directly

Required TodoWrite Items

  1. •file-analysis:root-identified
  2. •file-analysis:structure-mapped
  3. •file-analysis:patterns-detected
  4. •file-analysis:hotspots-noted

Mark each item as complete as you finish the corresponding step.

Step 1: Identify Root (file-analysis:root-identified)

  • •Confirm the analysis root directory with pwd.
  • •Note any monorepo boundaries, workspace roots, or subproject paths.
  • •Capture the project type (language, framework) from manifest files (package.json, Cargo.toml, pyproject.toml, etc.).

Step 2: Map Structure (file-analysis:structure-mapped)

  • •Run tree -L 2 -d or find . -type d -maxdepth 2 to capture the top-level directory layout.
  • •Identify standard directories: src/, lib/, tests/, docs/, scripts/, configs/.
  • •Note any non-standard organization patterns that may affect downstream analysis.

Step 3: Detect Patterns (file-analysis:patterns-detected)

  • •Use find . -name "*.ext" | wc -l to count files by extension.
  • •Identify dominant languages and their file distributions.
  • •Note configuration files, generated files, and vendored dependencies.
  • •Run wc -l $(find . -name "*.py" -o -name "*.rs" | head -20) to sample file sizes.

Step 4: Note Hotspots (file-analysis:hotspots-noted)

  • •Identify large files (potential "god objects"): find . -type f -exec wc -l {} + | sort -rn | head -10.
  • •Flag deeply nested directories that may indicate complexity.
  • •Note files with unusual naming conventions or placement.

Exit Criteria

  • •TodoWrite items are completed with concrete observations.
  • •Downstream workflows (architecture review, refactoring) have structural context.
  • •File counts, directory layout, and hotspots are documented for reference.

Troubleshooting

Common Issues

Command not found Ensure all dependencies are installed and in PATH

Permission errors Check file permissions and run with appropriate privileges

Unexpected behavior Enable verbose logging with --verbose flag