Adds first-class Stata execution and data-inspection capabilities to an MCP-enabled agent. The skill provides tools to run Stata syntax or .do files, load and page datasets, inspect variables and codebooks, retrieve stored r()/e() results, and export graphs for viewing. It surfaces return codes, stderr, and log paths so the agent can report precise diagnostics and incremental output for long-running jobs.
Use this skill whenever a user asks to run Stata commands, debug .do scripts, inspect dataset structure, request summaries or codebooks, export graphs, or retrieve stored results (r()/e()). It is especially valuable for econometrics workflows where reproducible command execution and access to Stata-native outputs matter.
Designed for MCP-capable agents and tooling that support remote tool invocation (Claude Code / MCP clients, Cursor-style MCP integrations).
Stata MCP skill provides comprehensive tool documentation for interacting with a local Stata MCP server — covering command execution, data loading/inspection, graph export, session management, and a UI data browser. The SKILL.md is well-structured with clear tool reference sections and practical examples, though it's essentially a detailed API reference rather than a skill with actionable automation scripts. Niche audience: Stata users who also use MCP clients.
No scripts bundled — pure documentation/reference skill. No security concerns: no shell injection, no credential exposure, no destructive commands, no network calls to suspicious hosts. The skill references localhost-only HTTP endpoints with bearer auth which is appropriate. Well-documented tool reference but very niche audience (Stata + MCP users). Architecture is decent with frontmatter and organized sections, but it's essentially one large monolithic SKILL.md without scripts/ or references/ directories.
