cursor-ide-stinger

Configures Cursor IDE hooks, rules, MCP registration, and extension wiring for Hivemind.

84|37|Updated May 23, 2026
One-click install
npx skills add https://github.com/legioncodeinc/vibe-coding-tools --skill cursor-ide-stinger-legioncodeinc
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: cursor-ide-stinger
Source: https://github.com/legioncodeinc/vibe-coding-tools/tree/main/src/skills/cursor-ide-stinger
Command: npx skills add https://github.com/legioncodeinc/vibe-coding-tools --skill cursor-ide-stinger-legioncodeinc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Cursor IDE integration involves several distinct surfaces with subtle, easy-to-break details: the Cursor 1.7+ hooks.json schema differs from Claude Code and Codex, .cursor/rules/.mdc files have four activation modes controlled by frontmatter, MCP server registration requires correct mcp.json entries, and the first-party extension has its own build and contribution model. This Skill consolidates authoritative guidance for all of these so changes to the Cursor surface stay correct, idempotent, and consistent. ## Core Features & Use Cases - Cursor hooks lifecycle wiring: Documents the six lifecycle events (sessionStart, beforeSubmitPrompt, preToolUse, postToolUse, afterAgentResponse, stop/sessionEnd), the Cursor-specific hooks.json shape, and the idempotent, Windows-safe merge logic in install-cursor.ts. - Rule file authoring: Explains .cursor/rules/.mdc frontmatter (description, globs, alwaysApply), the four activation modes, glob syntax, and anti-patterns such as overusing alwaysApply or reintroducing .cursorrules. - MCP and extension integration: Covers registering the Hivemind stdio MCP server in project or global mcp.json, plus the extension's manifest contributions, webpack build, and activation flow. - Use Case: When asked to add a new Cursor hook or fix a rule that is not firing, load this Skill to check the exact hooks.json entry shape and the rule's activation-mode frontmatter before editing. ## Quick Start Ask the agent to wire the Cursor hooks for Hivemind or to author a new .cursor/rules/*.mdc rule file, and it will consult the matching guide before making changes.

Frequently Asked Questions about cursor-ide-stinger

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I wire Cursor 1.7 hooks in hooks.json?▼

Add command objects directly under each event in ~/.cursor/hooks.json with type, command, and timeout fields, with no outer hooks wrapper per entry. The install-cursor.ts installer merges entries idempotently and only rewrites the file when content changes.

How do Cursor .mdc rule files activate?▼

Cursor rules in .cursor/rules/*.mdc activate in four modes based on frontmatter: alwaysApply true fires on every context, globs scope to matching files, description alone enables intelligent activation, and no fields means manual @-mention only.

How do I register an MCP server in Cursor?▼

Add a mcpServers entry to .cursor/mcp.json for project scope or ~/.cursor/mcp.json globally, pointing at the stdio server command, then restart Cursor since MCP config does not hot-reload. Keep secrets out of the file; the server reads its own credentials.

What is the difference between Cursor hooks.json and Claude Code hooks?▼

Cursor's hooks.json array entries are command objects directly with type, command, and timeout fields, with no outer hooks wrapper and no top-level matcher wrapper. A matcher such as Shell is a sibling field on the command object, used only on preToolUse.

Why does reinstalling Cursor hooks duplicate entries on Windows?▼

Duplication happens when matching prior entries without normalizing backslash paths. The installer normalizes paths to forward slashes before checking for the /.cursor/hivemind/bundle/ marker, so re-installs strip old entries instead of stacking duplicates.

When should I not use this Cursor IDE guidance?▼

Do not use it for TypeScript code quality of the installer or extension source, MCP protocol internals like tool schemas and transport, or harness wiring for Claude Code, Codex, or other agents. Those belong to separate specialist skills.