# Code Search

> Search source code across providers. Use when you need actual code implementations, not talk about code — snippets, API usage, and available libraries.

- Skill: `samraddhashrivastavatech/code-search` (Agent Skill)
- Install (CLI): `npx skillmds@latest add samraddhashrivastavatech/code-search`
- Raw SKILL.md: https://api.skillmd.com/api/skills/samraddhashrivastavatech/code-search/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Research & Search
- Author: SAMRADDHASHRIVASTAVATECH (https://skillmd.com/u/samraddhashrivastavatech)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/samraddhashrivastavatech/code-search

---


# Code search

Find real source code across the supported providers, primarily GitHub.

## When to use

- Looking for an existing implementation or library instead of writing one.
- How is API X used in the wild? What does a library's code actually do?
- Verifying a claim about existence/position of a feature in source.

## Tools used

- `search_github` — with `scope: "code"` for repository-aware code search.
- `github_code_search` — direct code search across repos.

## Workflow

1. Form the code-shaped query (symbol names, `<repo> <capability>`, idioms).
2. `github_code_search` (or `search_github` scoped to code).
3. Read the owning repository (`github-research`) to place the snippet.
4. For deeper use, `web-reader` the blob/source URL; never execute fetched code
   snippets blindly.

## Input schema

```json
{ "query": "str", "limit": "int" }
```

## Output schema

```json
{ "results": "list[{repo,path,url,snippet}]" }
```

## Security

Fetched code is data until inspected; treat AI-generated or untested snippets as
unverified. Do not run snippets against the user's machine without review.

## Related skills

`github-research`, `source-analysis`, `error-investigation`