Unit 4: Prompt injection: detect and contain · Lesson 5 of 5 · 18 min Pro
Case file: an agent session that crossed repositories
Trace an agent session that read an untrusted issue, then private repos, then wrote publicly, and check the fixed token.
Unit 4: Prompt injection: detect and contain · Lesson 5 of 5 · 18 min Pro
Trace an agent session that read an untrusted issue, then private repos, then wrote publicly, and check the fixed token.
Unit 4: Prompt injection: detect and contain · Lesson 5 of 5 · 18 min Pro
Trace an agent session that read an untrusted issue, then private repos, then wrote publicly, and check the fixed token.
Pro opens every lesson in AI and LLM Security and the other courses, with unlimited lab machine time and certificates.