AI & cybersecurity, explored. Field guides, learning notes, and experiments.https://promptexploit.com/en-ushttps://promptexploit.com/journal/when-retrieved-text-becomes-an-instruction/https://promptexploit.com/journal/when-retrieved-text-becomes-an-instruction/A closer look at indirect prompt injection, the RAG trust boundary, and why retrieved content should never inherit authority.Sat, 12 Sep 2026 00:00:00 GMThttps://promptexploit.com/journal/authorization-is-more-than-a-login/https://promptexploit.com/journal/authorization-is-more-than-a-login/Follow one request from identity to resource access, and find the security decision that authentication alone cannot make.Sat, 12 Sep 2026 00:00:00 GMThttps://promptexploit.com/journal/before-the-exploit-draw-the-boundary/https://promptexploit.com/journal/before-the-exploit-draw-the-boundary/A practical starting point for threat modeling: identify the assets, follow the data, and ask where trust changes.Sat, 12 Sep 2026 00:00:00 GMThttps://promptexploit.com/journal/an-agent-tool-is-a-permission-boundary/https://promptexploit.com/journal/an-agent-tool-is-a-permission-boundary/Look past the tool description. What can the agent actually do, with whose identity, and under which constraints?Sat, 12 Sep 2026 00:00:00 GMThttps://promptexploit.com/journal/a-lab-note-you-can-reproduce/https://promptexploit.com/journal/a-lab-note-you-can-reproduce/A small template for turning an interesting result into a useful experiment: setup, observation, explanation, and limits.Sat, 12 Sep 2026 00:00:00 GMThttps://promptexploit.com/journal/llm01-2025-prompt-injection-simple-explanation/https://promptexploit.com/posts/llm01-2025-prompt-injection-simple-explanation/Understand direct and indirect prompt injection through a workplace example, practical defenses, and a safe trust-boundary exercise.Thu, 04 Jun 2026 00:00:00 GMThttps://promptexploit.com/journal/llm02-2025-sensitive-information-disclosure-simple-explanation/https://promptexploit.com/posts/llm02-2025-sensitive-information-disclosure-simple-explanation/Follow private data through an AI assistant, distinguish training from retrieval, and practice checking who may see each record.Thu, 04 Jun 2026 00:00:00 GMThttps://promptexploit.com/journal/llm03-2025-supply-chain-simple-explanation/https://promptexploit.com/posts/llm03-2025-supply-chain-simple-explanation/Learn how models, packages, adapters, and providers enter an AI system, and how to review their origin before trusting them.Thu, 04 Jun 2026 00:00:00 GMThttps://promptexploit.com/journal/llm04-2025-data-and-model-poisoning-simple-explanation/https://promptexploit.com/posts/llm04-2025-data-and-model-poisoning-simple-explanation/Distinguish poisoned training data, tampered models, and corrupted retrieval sources, then practice tracing a false policy back to its origin.Thu, 04 Jun 2026 00:00:00 GMThttps://promptexploit.com/journal/owasp-genai-llm-top-10-simple-explanation/https://promptexploit.com/posts/owasp-genai-llm-top-10-simple-explanation/Understand all ten OWASP LLM risks in the 2025 edition through plain-language examples, first defenses, and a small design exercise.Fri, 29 May 2026 00:00:00 GMThttps://promptexploit.com/journal/indirect-prompt-injection-101/https://promptexploit.com/posts/indirect-prompt-injection-101/Follow a harmless-looking document into an AI assistant and learn where outside content can become an unauthorized instruction.Fri, 15 May 2026 00:00:00 GMThttps://promptexploit.com/journal/isolating-tool-output-in-agents/https://promptexploit.com/posts/isolating-tool-output-in-agents/Keep information from tools separate from permission to act, with clear examples of provenance, narrow tools, and independent authorization.Mon, 20 Apr 2026 00:00:00 GMThttps://promptexploit.com/journal/building-a-jailbreak-eval-harness/https://promptexploit.com/posts/building-a-jailbreak-eval-harness/Build a small, repeatable security evaluation that checks real outcomes, catches regressions, and explains what its scores mean.Tue, 10 Mar 2026 00:00:00 GMT