Pratip Dasgupta·Jun 3AgentSigma v0.1 — Building the Detection Framework I Wrote AboutPart 1 made the argument. This is the first working proof.
Pratip Dasgupta·May 11Detection Engineering Is the Missing Discipline in AI Security — Part 1The alert that never fired, the three tribes that missed it, and why your SIEM cannot see the attackA response icon1A response icon1
Pratip Dasgupta·Oct 20, 2025Beyond Context — Treating Prompt Injections Like MalwareBuilding Signature + Heuristic Defenses for LLMs
Pratip Dasgupta·Oct 13, 2025🧠 Memory Makes Prompt Injections DangerousPart 3 of my LLM safety experiments — bringing context awareness to Inj3ctStop.
Pratip Dasgupta·Sep 29, 2025Beyond Naïve Filters: Chaining Guardrails for LLM Prompt InjectionNext step of my LLM safety experiments — moving from simple regex filters to chained guardrails
Pratip Dasgupta·Sep 21, 2025When LLMs Obey Too Much: Playing with Prompt Injection DefensesThis trick is called a prompt injection, and it’s one of the biggest headaches for anyone building LLM-powered apps today.
Pratip Dasgupta·Sep 17, 2025Breaking AI to make IT safer“You don’t know how strong a lock is until someone actually tries to pick it.”
Pratip Dasgupta·Sep 13, 2025“Security teams spend more time sorting vulnerabilities than fixing them.”That sentence has been bouncing around in my head for months.