If an agent can still move money after its prompt says no, the boundary is in the wrong place
ExecLayer_io · reddit · 2026-11-22
ExecLayer's founder argues that agent safety must live in tools and credentials, not prompts. Reading email, accessing customer records, paying, and changing bank details are separate powers; payment authority should be a distinct grant. The real danger of prompt injection is untrusted content influencing privileged actions, so every request should be checked against rules outside the model — e.g. an agent that may read invoices but requests a payee change gets refused, and actions don't run when required checks can't complete. He also stresses testing denied paths (extra API keys, direct network calls) and keeping an audit trail linking actor, action, rule, and decision, noting vendor "safe model" claims don't substitute for per-action authorization.
More from coding & agent
- DeepORG open-sources an AI agent that runs across your org's databases, APIs and code — dosco · 2026-11-22
- AutoRAGBench: LLM agents build RAG systems beating the best published human designs — omarsar0 · 2026-11-22
- Stop @-ing your agent in Slack — let Jev decide when it should respond — DavidWells · 2026-11-22
- Intent adds notch launcher, Claude Code and Codex session import — Wattenberger · 2026-11-22
- rakyll explains AX: a low-opinion job orchestrator that agents control themselves — rakyll · 2026-11-22
- Agent builds a full Shopify storefront with zero manual edits under a code grammar — FrBatista · 2026-11-22