grepai-mcp-tools_skill

This skill helps you understand GrepAI MCP tools, their parameters, and usage to integrate, debug, and optimize code search and call-graph tasks.

1

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill yoanbernabeu/grepai-skills --skill grepai-mcp-tools

SKILL.md8.3 KB

Overview

This skill is a concise reference for all GrepAI MCP tools. It helps you discover available tools, understand their parameters, and choose compact vs full output for efficient integration with AI assistants. Use it to inspect call graphs, trace callers/callees, run semantic code search, and check index health.

How this skill works

The skill lists each MCP tool, its required and optional parameters, example requests, and both compact and full response shapes. It explains compact key mappings to reduce token usage and shows common error responses and MCP request/response envelopes. Use the examples to form valid JSON-RPC requests to the MCP server.

When to use it

When you need semantic code search across a codebase
When tracing which functions call or are called by a symbol
When building a call graph starting from a root function
Before running searches to verify index health and embedder connectivity
When integrating GrepAI tools into an AI assistant or automation pipeline

Best practices

Prefer compact mode to reduce token usage (~80% savings)
Limit result counts to only what you need
Check grepai_index_status before heavy queries
Combine search with trace tools for full context (search -> trace_callers/trace_callees)
Handle and surface error codes (INDEX_NOT_FOUND, EMBEDDER_UNAVAILABLE, SYMBOL_NOT_FOUND)

Example use cases

Find all occurrences and code snippets related to "authentication middleware" with grepai_search
Trace who calls Login across the codebase with grepai_trace_callers
List all functions invoked by ProcessOrder using grepai_trace_callees
Build a multi-level call graph from main to visualize execution paths with grepai_trace_graph
Verify index and embedder status before running large semantic queries with grepai_index_status

FAQ

Compact mode replaces full keys with abbreviated keys (e.g., query->q, results->r) and trims content to reduce token size by roughly 80%.

How do I call a tool via MCP?

Send a JSON-RPC tools/call request with name and arguments. The skill shows example request and the expected result envelope.

What should I do if the index is missing?

Run the project indexing workflow (e.g., start watching or build the index) before calling search or trace tools; check grepai_index_status for diagnostics.