🔐 Auth & security
3 skills and 2 prompts. Each page has a prompt you can paste into any agent, and skills install with one command.
← all 120 skills and prompts- Auth & permissions done rightexample
Sign-in with Better Auth or Supabase Auth, role checks in one module, and RLS backing every query.
- Build my redactorskill
Builds a one-file redaction tool that runs only in your browser. It swaps emails, phone numbers, serial numbers, asset tags, addresses and names you list for placeholders before you paste anything into AI. It adds a pre-prompt that tells the AI to keep the placeholders, and a restore button that puts the real values back in the answer. Use when someone wants to use AI with customer or company data without that data leaving their device.
- MCP Dependency Reviewskill
Reads a workspace's MCP config files as text and flags server packages and images that can change under an approved config (bare names, @latest and other tags, version ranges, Docker images without a digest), classifies each, and says what to pin, without running anything. Use before approving a new or changed MCP config, when adding an MCP config check to CI, or when a server launches through npx, bunx, pnpm dlx, yarn dlx, uvx, pipx or docker run.
- Prompt-injection defense for agentsexample
Harden a tool-using agent against indirect prompt injection with capability limits, approvals, red-team CI.
- Redact with buttonsskill
Builds a one-file redaction tool with a big button for each kind of detail to hide (serial numbers, asset tags, emails, phones, IP and MAC addresses, street addresses, prices, dates, PO numbers, names, "made in"). Tap the ones you want, paste text, and it swaps them for placeholders before you ask AI, then puts the real values back in the answer. Runs only in your browser, even in airplane mode. Use when someone wants to use AI with customer or company data without that data leaving their device.