AI Skill MartBrowse skillsPromptsProgramsLearnWhat is a Skill?
 
← Back to list

Smart Explore

Token-efficient codebase navigation that uses tree-sitter AST parsing to map structure first and only unfold the symbols you actually need.

Dev & CodingIntermediate★ 90,139⑂ 7,846AI score 9/10Last updated: Aug 9, 2026

What it does

  • Enforces a 3-layer exploration workflow: smart_search (find files + ranked symbols across a directory) → smart_outline (structural skeleton of one file) → smart_unfold (full source of one symbol).
  • Replaces the Glob → Grep → Read discovery loop with a single structural search call.
  • Includes a token-economics table so you know when each approach is cheapest (full Read ~12k tokens vs outline+unfold ~3k).
  • Explains exactly when to fall back to standard tools: Grep for exact regex, Read for small/non-code files, Explore agent for cross-file architectural synthesis.
  • Bundled tree-sitter grammars for JS/TS/TSX, Python, Go, Rust, Ruby, Java, C, C++, plus custom grammar registration via .claude-mem.json and special Markdown heading support.

Who it's for

  • Developers working in large repos where context windows fill up fast.
  • Teams optimizing token spend without losing code-reading accuracy.
  • Users of the claude-mem plugin, which supplies the required MCP tools.

Example uses

  1. Trace a cross-cutting feature: smart_search(query="shutdown", path="./src") surfaces 14 symbols in 7 files, then unfold only performGracefulShutdown.
  2. Navigate a 1,400-line service file: outline it (~1.5k tokens), then unfold the one method you need (~1.6k) instead of a 12k-token full read.
  3. Write docs about a feature: smart_* tools for code discovery, plain Read for small config/markdown/plan files.

Prerequisite: smart_search / smart_outline / smart_unfold are MCP tools from claude-mem. Without that server connected, the skill has nothing to call.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/thedotmack/claude-mem/HEAD/plugin/skills/smart-explore/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on AI Skill Mart.
Copy the plugin/skills/smart-explore folder from the GitHub repo thedotmack/claude-mem into my ~/.claude/skills/smart-explore/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/thedotmack/claude-mem.git && mkdir -p ~/.claude/skills && cp -r claude-mem/plugin/skills/smart-explore ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Clone the repository: git clone https://github.com/thedotmack/claude-mem.git
  2. Copy the skill into your personal skills folder: mkdir -p ~/.claude/skills && cp -r claude-mem/plugin/skills/smart-explore ~/.claude/skills/
  3. Install and register the claude-mem plugin / MCP server following the repo README — this is required, since the skill only contains instructions.
  4. Restart Claude Code and verify with /mcp (or the tool list) that smart_search, smart_outline, and smart_unfold are available.
  5. Test it inside a project: ask "use smart-explore to find shutdown-related code in ./src".
  6. Optional: for languages outside the bundled list, add a .claude-mem.json at your project root with package, extensions, and a query file, then run npm install tree-sitter-<language>.
View source on GitHub ↗License: Apache-2.0