A vault per conversation and per job: a leaked key exposes one small thing, not an inbox

Working this way you start making very small vaults for very specific things: a communication vault for the two or three agents on a job, in which they talk and work through versioned, encrypted files. What an agent needs to join one is small: the vault key, sgit and network access to one host, or a server inside its own network. This is the right shape for security: a leaked key exposes one small vault and one conversation, not an inbox, a drive or an account, and every word is on the record for the person who holds the keys. It is also why the number of keys grew faster than one person could manage by hand.

Decided by. Nobody yet: a candidate for Dinis to read, comment on and decide (keep, rewrite, merge, cut, article first, later, not doing it).

Note. Pairs with an-agent-is-a-session-with-a-focus (one session per topic) and with the main workflow the article describes: give Claude no permissions and no access to anything, only the vault it uses to communicate, or a vault it creates and the person then captures. Inside the vaults the agents use EmailFS and IssuesFS, file-based protocols for mail and issues.

Where it comes from

Where the vault keys live Small vaults for small jobs (newsroom)

What it answers in the 2018 book

none named

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: Mandate, reach and the gap: hope versus enforcement, 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, The agent behaviour policy is the due diligence document: a buyer reads the blast radius off it, and its absence says the most, The desktop is the barrier, not the model: prompt injection is not solved, so the box has to hold, Reach is a computed value, and new capability lands as excess, never as new authority, Who can pull the plug: an agent that cannot be stopped is riskier than one that can, A third input: a behaviour policy beside the system prompt and the prompt, which the vendor acknowledges, enforces where it can and answers for, Models do not separate code from data, so a rule in the prompt can be argued with by an email, Aviation and payments got safe when the rule, the liability and the business model pointed the same way, Zoom into an agent's behaviour policy and you find the business logic, which only the business can write, Software was the law: the rule the screen enforced by omission is no longer enforced once an agent calls the API, Every rule lands in one of three states: backed by a control, governed by an accepted risk, or resting on hope, Write down what your agent can reach in twenty minutes: list its tools, say what you wanted it for, write the gap.

Node idea:small-vaults-for-small-jobs in graph.json; see it in the graph. Source: admin/content/graph/ideas.json.