agentsclimarketplace

Researching codebase

Skill qte77/claude-code-plugins/plugins/codebase-tools/skills/researching-codebase

A Claude Code plugin marketplace providing skills, rules, and scripts extracted from a production development workflow.

Install
npx -y skills add qte77/claude-code-plugins --skill researching-codebase

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Investigates codebase before planning. Use before any non-trivial implementation task to gather context in isolation.

SKILL.md

2.4 KB, as published. Nobody here has run it

Codebase Research (ACE-FCA)

Query: $ARGUMENTS

Gathers codebase context in isolation before planning. Prevents search artifacts from polluting main context.

Core Principles

  1. Documentation-only - Describe what exists, where, and how it works
  2. No evaluations - Never suggest improvements or critique implementation
  3. Evidence-based - Provide file paths, line numbers, and code references
  4. Isolation - Research runs in fork context; return only distilled findings

When to Use

  • Before planning non-trivial implementations
  • When unfamiliar with relevant codebase areas
  • Before architectural decisions

Workflow

  1. Read mentioned files first - If specific files mentioned, read completely before exploring
  2. Decompose the question - Break query into researchable components
  3. Explore codebase - Investigate architecture, patterns, constraints
  4. Identify scope - Determine relevant areas based on findings
  5. Distill - Return structured summary using output format below

Output Format

Follow ACE-FCA quality equation: Correct + Complete + Minimal noise

---
research_query: "<original question>"
timestamp: "<ISO 8601>"
files_examined: <count>
---

## Key Files

| File | Purpose | Key Lines |
|------|---------|-----------|
| `path/to/file.ext` | Brief purpose | L42-58 |

## Patterns

- **Pattern name**: Description with file reference (`path:line`)

## Constraints

- Constraint with evidence (`path:line`)

Evidence Requirements

Every claim must include:

  • File path: Exact location (src/auth/login.ts)
  • Line numbers: Specific lines (L42-58 or L127)
  • Code reference: Function/class name when relevant

Bad: "Authentication uses JWT tokens" Good: "Authentication uses JWT tokens (src/auth/jwt.ts:L23-45, verifyToken function)"

References

See references/context-management.md and references/core-principles.md.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.