-rw-r--r-- 10m Agent runtime An unknown outcome is its own state
A process crashes mid-execution. The audit log shows the action was authorized. The runtime logs show it was attempted.
authorization
root@amaltsev >:~/blog$ grep -l "authorization" -r ~/blog
16 posts on authorization.
root@amaltsev >:~/blog$ grep -l "authorization" -r ~/blog
mode date min topic name
-rw-r--r-- 10m Agent runtime Four valid permissions and one chain nobody approved
Every step is in scope. The chain is a bulk data export to an external recipient. The agent executes them in order, because the task it was given was "produce a customer report".
-rw-r--r-- 10m Agent runtime Operator control lives in state the agent cannot reach
The worker is on another pod. The queue keeps delivering jobs. Another worker picks up the in-flight execution.
-rw-r--r-- 9m Agent runtime Agents cannot widen their own authority
There is a question most governance frameworks never ask, because it does not come up until an agent can write code, configure integrations or modify its own workflows.
-rw-r--r-- 9m Agent runtime Authority is durable, and evidence expires on its own clock
An agent is authorized to issue refunds up to five hundred dollars. The permission is valid, and the agent uses it correctly.
-rw-r--r-- 10m Agent runtime Scope an agent's permissions to the credential it holds
Three agents, one platform. The support agent reads the CRM, opens tickets, and refunds up to fifty dollars.
-rw-r--r-- 12m Agent runtime Reducing approval fatigue without removing the control
A team wants an agent that issues refunds. Security is nervous. There are exactly two options on the table, and both are bad.
-rw-r--r-- 15m Agent runtime Which fields are material? What makes a digest mean something
A tool is described by a JSON schema: its parameters, their types, what it accepts. This vocabulary fits an interface.
-rw-r--r-- 8m Agent runtime An approval is durable security state
The sequence is short. An agent submits an action for approval. The operator approves it. The agent crashes.
-rw-r--r-- 10m Agent runtime Binding a human approval to the exact payload it authorized
Every framework ships a human-in-the-loop step. A human sees a card, the human clicks approve, the tool runs.
-rw-r--r-- 11m Agent runtime How human-in-the-loop approvals get bypassed on agents
A team ships a human-in-the-loop approval step. The tests pass, the demo works, the security review signs off.
-rw-r--r-- 11m Agent runtime Six enforcement paths, six declared defaults
Fail-closed is a one-sentence rule: when the rule cannot be evaluated, refuse. Every enforcement system claims it.
-rw-r--r-- 8m Agent runtime An agent budget is an authorization decision
The first version of this always looks the same. There is a number, the number grows, and when the number crosses a threshold the request is rejected.
-rw-r--r-- 10m Agent runtime Every authorization is an instance of a permission category
Same agent. Same tool. Same permission. A support agent with refund access approves a twenty-dollar refund, which is exactly what it is there for.
-rw-r--r-- 9m Agent runtime Intercepting tool calls without patching framework internals
The request that shows up repeatedly in agent framework issue trackers is a developer who wants to inspect and block a tool call before it executes.
-rw-r--r-- 8m Agent runtime Treat an agent as two systems joined by a seam
An agent produces a perfectly acceptable response and then sends the wrong payment, deletes the wrong resource, modifies the wrong record, or calls a tool with a mutated argument.
~/blog: 16 matches
root@amaltsev >:~/blog$ cd ~/blog && ls -la
The whole directory is one click away. Every post, filed by subject.