ConeheadAI · leftover lesson

SNK3-025 · Curriculum How-to

Simon prompt injection article — high value handoff

L1–L3 · Hands On–Verifier (bootstrap — system rebalances) · ~5m

harnessprompt_injectionVER · J4 · S2C1.4C2.3

Look for this insight

Simon’s injection write-up: untrusted content can contain instructions. True/false that risk.

Lesson hook

When an outside resource is genuinely excellent, we'll send you straight to it. Simon Willison's writing on prompt injection is well worth the click — just read the definition plus one example (5 minutes, tops).

Do this now

Click the snack and read until you can confidently answer: true or false — prompt injection can hide inside documents the model is asked to process. (Spoiler for after you've read it: true.)

Completion check: true_false

Open snack link →

Staff review Sign in

Related by competency

All snacks · Take Question 1 · Team training