Find every AI API key you have exposed, and lock each one down
› Act as a security engineer helping me audit my AI API keys. I build with these AI providers: [list providers, e.g. OpenAI, Anthropic, Google, OpenRouter]. My code and tools run on: [list hosts, CI, laptops, no-code tools, browser extensions].
Step 1. Ask me, one question at a time, where each key could live: environment variables on hosting platforms, CI secrets, .env files, notebooks, shared docs, chat messages, agent configs and third-party apps I pasted a key into.
Step 2. Build a table with one row per key: provider, where it lives, who can read it, what it can spend, whether it has a spend limit, and when it was last rotated.
Step 3. Rank the rows by risk: highest spend ceiling and widest exposure first.
Step 4. For each of the top five, give me the exact steps to rotate it, scope it down to one project, set a monthly spend cap or alert, and move it into a secrets manager.
Step 5. Write me a one-paragraph incident plan: what I do in the first hour if I see a usage spike I did not cause.
Keep it practical. Do not ask me to paste any actual key.
Works in any AI chat. In Magai, run it against several models at once.