Study

What it does

When your agent uses it

• Asks to study this codebase
• Extract patterns from this project
• Create a blueprint
• Learn the architecture for building my own version

Example prompts

• “study this codebase”
• “extract patterns from this project”
• “create a blueprint”
• “/study”

Requirements

• Node.js
• Compatibility (from SKILL.md): Requires grepika MCP server.
• Pre-approved tools (allowed-tools): mcp__grepika__search, mcp__grepika__refs, mcp__grepika__outline, mcp__grepika__toc, mcp__grepika__stats, mcp__grepika__get, mcp__grepika__context, mcp__grepika__graph, mcp__grepika__structural_search, mcp__grepika__add_workspace, Bash

Workflow steps

6 steps, taken from the step headings in SKILL.md.

1. 1Detect Source Name
2. 2Get Codebase Overview
3. 3Identify Key Architectural Patterns
4. 4Assess Dependencies
5. 5Determine Build Order
6. 6Write Blueprint File

What it can do on your machine

Read from SKILL.md and the folder at commit f035a18. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves these tools, so the agent can use them without asking each time:

  • mcp__grepika__search
  • mcp__grepika__refs
  • mcp__grepika__outline
  • mcp__grepika__toc
  • mcp__grepika__stats
  • mcp__grepika__get
  • mcp__grepika__context
  • mcp__grepika__graph
  • mcp__grepika__structural_search
  • mcp__grepika__add_workspace

  …and 1 more on the same allowed-tools line.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • jq
  • git

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

• Compatibility

  Requires grepika MCP server.

  From compatibility in the SKILL.md frontmatter.

Context cost

Study loads about 1.6k tokens when it runs. Until then it costs about 85 tokens; SKILL.md has 461 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~85

When it runs · the whole SKILL.md, loaded when a task matches

~1.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

• NotePre-approves every shell command (allowed-tools: Bash)SKILL.md

  allowed-tools: mcp__grepika__search, mcp__grepika__refs, mcp__grepika__outline, mcp__grepika__toc, mcp__grepika__st

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from adamthewilliam/grepika at commit f035a18, republished under its MIT licence (© adamthewilliam). 461 words, ~1,605 tokens.

name

study

description

This skill should be used when the user asks to "study this codebase", "extract patterns from this project", "create a blueprint", "learn the architecture for building my own version", "what patterns can I reuse from this codebase", or mentions keywords related to extracting reusable architectural patterns for use in other projects.

allowed-tools

mcp__grepika__search, mcp__grepika__refs, mcp__grepika__outline, mcp__grepika__toc, mcp__grepika__stats, mcp__grepika__get, mcp__grepika__context, mcp__grepika__graph, mcp__grepika__structural_search, mcp__grepika__add_workspace, Bash

compatibility

Requires grepika MCP server.

version

1.0.0

disable-model-invocation

true

context

fork

agent

Explore

model

sonnet

mkdir -p ~/.claude/blueprints/SOURCE_NAME
cat > ~/.claude/blueprints/SOURCE_NAME/blueprint.md << 'BLUEPRINT_EOF'
[blueprint content here]
BLUEPRINT_EOF

---
source: [SOURCE_NAME]
path: [absolute path to source codebase]
studied: [YYYY-MM-DD]
patterns: [count]
focus: [focus area or "full"]
---

# Blueprint: [SOURCE_NAME]

> [One-sentence description of what this codebase does]

## Overview

- **Language**: [primary language]
- **Framework**: [key framework/runtime]
- **Architecture**: [brief architecture style description]
- **Key Dependencies**: [3-5 most important dependencies with purpose]

## Build Order

1. [pattern-name] — [why first]
2. [pattern-name] — [builds on #1]
3. [pattern-name] — [builds on #1-2]
...

---

## Pattern: [kebab-case-name]

**Load-bearing**: [yes/no]
**Category**: [category]

### What
[One paragraph describing the pattern]

### Why
[The constraint or problem that motivated this choice]

### Key Files
- `path/to/file.rs:42` — [what this location shows]
- `path/to/other.rs:100` — [what this location shows]

### Implementation
```[language]
[Essential code snippets — brief, focused on the pattern]

## Output

After writing the blueprint, report:

## Guidelines

- **Be selective**: Extract 4-10 patterns, not every function. Focus on architectural decisions, not implementation details.
- **Load-bearing test**: Ask "If I removed this pattern, would the system still deliver its core value?" If yes, it's not load-bearing.
- **Why > What**: The rationale is more valuable than the code. Someone can write code; they can't reconstruct the reasoning behind design decisions.
- **Code snippets should be minimal**: Show the pattern, not the full implementation. 10-30 lines per pattern.
- **Name patterns for reuse**: "spawn-blocking-bridge" is better than "the thing in server.rs". Names should make sense outside the source project.

© adamthewilliam, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in plugins/grepika/skills/study of adamthewilliam/grepika.

Open the folder on GitHubat commit f035a18