frame candidate goes to: this-edition T town planner
Mandate, reach and the gap: hope versus enforcement
Before you give an agent access, write down what it is for and what it can reach; the gap is the risk; rules in a prompt are hope, rules in a hook or platform are enforcement; behaviour, decision and workflow policies are different things.
Decided by. Nobody yet: a candidate for Dinis to read, comment on and decide (keep, rewrite, merge, cut, article first, later, not doing it).
Where it comes from
- What can your agent reach? RiskMandate and the Agent Behaviour Policy, explained for the person who has to decide the plain explainer (newsroom)
- Are hooks the sweet spot? Moving an agent's rules from hope into hooks, rule by rule, until enforcement moves off the machine hope-based security (newsroom)
- Attacks rewritten as mandates 28 incidents (newsroom)
- Three policies for an agent the three-policy split (newsroom)
What it answers in the 2018 book
- Threat Modeling answered outline, Security
- 3rd-party-modules answered outline, Security
Where it goes
Chapter Controlling your agent: RiskMandate, part Build it without a server (drafting); topic Controlling your agent: mandate, reach and hope versus enforcement.
Other ideas proposed for the same chapter: Code is a liability; the future is the smallest code that does the job, Learning to hack is still how you learn how things work: draw the line where the attacker is, The secret goes from the drop into the tool, never into the chat, Security experience is what lets you ship with known gaps: list them, match them to the reader, say what would change the call.
Node idea:mandate-reach-gap in graph.json; see it in the graph. Source: admin/content/graph/ideas.json.