Skip to content
A personal experiment · alpha · code not public yet · no offer, no support Project status →

Agent sessions and sandboxes

A coding agent works freely in a separate place, and the one step that changes production goes through an approval. For that to hold, the separate place must really be separate: its own credentials, no production secrets, no network path to production, and disposable.

A trigger (a webhook, a node signal, a schedule) or a human starts an agent session. The core decides when, whether and with which rights; the driver only starts, polls and stops it.

Where the session runsDriverHow the session proves itself
a pod in kubernetes-sigs/agent-sandbox on k3sagent_sandboxa projected ServiceAccount token, checked by TokenReview; bound to namespace, ServiceAccount and pod
a Forgejo Actions jobforgejo_actionsthe job’s OIDC ID token (ES256), bound to repository and run
your machine, next to youvaicoa-run paira comparison code you confirm on your phone
  • The session gets no token, but a run ticket: usable once, valid for at most ten minutes. Redeeming it gives a run certificate that ends with the run’s budget.
  • The sandbox may reach only the core’s door and the model (a NetworkPolicy). On Forgejo runners the driver cannot restrict egress (limitations).
  • Every run has a budget: tokens, runs per day, proposals per run, a timeout.
  • A run carries at most the rights of the user it acts for, capped at level device, and only the actions in its trigger’s mayPropose list.

The agent never talks to the core directly. vaicoa-run sits in the session with two faces: a CLI and a local MCP server over stdio, for Claude Code, OpenCode or any MCP client. Its tools are the allowed actions from the verified signed catalog, plus:

ToolWhat it does
actionswhat this run may propose, and for each action whether it runs at once or needs approval, and from whom
explaina dry run: what the core would decide for exactly these parameters, without creating an op
proposeanswers done, queued, pending_approval (with what is missing) or refused (with the reason)
status, waitfollow an op, within the run’s budget
reportsend results or questions back to the human

A skill file tells the agent how to behave: explain first, propose once, keep working while an approval is pending, do not repeat a proposal. Reviewer sessions get review-next and review-decide instead and never see the proposer’s claims. A run never has both roles.

The idea behind the name: an agent gets a copy of an environment (with seed data or a masked copy), tries a change there and presents the result for review. This is not built in Vaicoa Core. What exists is the session in a sandbox or CI job, changes delivered as a pull request (GitOps flow), and plans that show before and after.