Explore

Best practices for composing multiple rulesets?

I have three .cursorrules files. What is the recommended way to layer them?

How many AI tools do you use daily?

Gauging the multi-tool reality of AI-assisted development.

How to handle .cursorrules conflicts in monorepos?

We have a Turborepo with 12 packages. Each team wants their own .cursorrules. Currently we concatenate at build time but

Best practices for MCP server caching

The filesystem MCP server re-reads on every tool call. Has anyone added a read-through cache?

Devin-style rules + Cursor Composer = magic

The AI now plans before coding and catches its own mistakes.

Which AI coding environment is your primary tool?

We want to understand the community's tool preferences to better prioritize environment support.

MCP server authentication with OAuth2?

I need my MCP server to authenticate against our corporate IDP (Okta). The MCP spec doesn't seem to cover auth flows. Ha

Automated PR review pipeline with Claude Code

Full CI pipeline: push to GitHub, Claude reads the diff via GitHub MCP, posts review comments, triggers fixes, re-review

Screenshot quality tips

Default screenshots are 1280x720. Pass {width: 1920, height: 1080, deviceScaleFactor: 2} for retina.

Windsurf rules vs Cursor rules — are they interchangeable?

Can I use the same .cursorrules file in Windsurf or do I need to adapt? The documentation is unclear on format differenc

Figma-to-production in 20 minutes

Design in Figma, Figma MCP reads frames, Claude generates Tailwind components, Playwright MCP verifies the render. The f

Python rulesets feel stale

The Python/Django subset references Django 4.x. Django 5.2 changed several idioms.

Would you pay for verified MCP servers?

Verified MCP servers would include security audits, uptime guarantees, and official support.

Claude Code skill vs system prompt — when to use which?

I have been writing system prompts for everything but just discovered Claude Code skills. What is the practical differen

Postgres MCP fails on large result sets

Queries returning >1000 rows cause the MCP server to OOM.

Testing AI-generated code quality — metrics that matter?

Beyond pass/fail tests, what metrics do you track for AI code quality? Looking for something more nuanced than 'it compi

Using GitHub MCP for automated PR reviews

I have Claude reading and commenting on PRs across 3 repos.

What category needs more content?

Help us decide where to focus community curation efforts.

Rate limiting strategies for Context7 in CI?

Our CI runs 200+ parallel jobs. Each calls Context7 for doc lookup. We are hitting rate limits. What caching layer works

n8n + Claude for customer support triage

Incoming support tickets via webhook, classified by Claude, routed to the right team in Linear, response drafted. Cut re

My Cursor setup for large monorepos

I keep a base .cursorrules at the repo root and override per-package via symlinks.

How to version control AI agent configurations?

Our CrewAI agents have configs that drift between environments. Git is awkward for JSON blobs. What do teams use for age

RIPER modes in team settings

Enforce specific RIPER modes per branch.

Best practices for prompt injection defense in production?

We ship a customer-facing AI feature. What are proven defenses against prompt injection beyond basic input sanitization?

Webhook trigger best practices

What is the recommended way to secure webhook triggers in n8n?

How do you discover new AI configs?

Understanding discovery channels helps us improve distribution.

TDD-first Rust development with Cursor

Using TDD skill + Cursor rules for a Rust CLI. Claude writes failing tests, implements until green, then refactors. 94%

Which prompts still work with GPT-5.4?

Some prompts were written for GPT-3.5. Has anyone tested with the latest?

Structured prompts vs free-form

Structured for complex tasks, free-form for creative work.

Preferred license for open-source rulesets?

Which license model makes you most comfortable sharing?