Topic tag
agent safety
5 guides in the catalogue cover this topic.
- Nº 010 Expertise you can install Agents & Automation · Broad
A skill file is career-long judgment packaged as instructions your agent can run. How to adopt someone's method, and why you scan it like untrusted input first.
- Nº 016 Guardrails for hours-long agent runs Prompting & Context · Deep
Guardrails are the support structure that keeps an hours-long agent run honest: an audit instruction, a memory that survives resets, and fresh-eyed verifiers.
- Nº 032 Defend your agent from prompt injection Agents & Automation · Practitioner
Anything your agent reads (a web page, a resume, a calendar invite) can carry hidden instructions it will follow. Why the attack has no clean fix, and the five defense layers that shrink what a successful one can do.
- Nº 036 Assume your agent uses every tool it has Agents & Automation · Practitioner
An agent sent a discount code to 150,000 people because a to-do item looked like an instruction. A rule in the prompt is a suggestion; a rule built into the tools is a restriction. Scope every key to the least the task needs.
- Nº 038 Delegate to your agent in stages Agents & Automation · Broad
A new agent should not run your inbox on day one. Give it one job with read-and-draft access only, check its work until your corrections stop, then add the next tool, and let it run unattended only after it has earned that.