Files
docs/working-with-the-agent.md
T
solo.ist 424cfbdf5e docs: add "Working with the agent" (norms for reading and instructing inferencebot)
A first draft of the member-facing norms listed as a design goal on
co-op/support#17: where the agent speaks and listens, what to ask it for,
what not to ask it in a shared room, what it does without confirmation,
what always needs a human, and how it is stopped. Linked from the README
index and the Matrix section of getting-started. Nothing here changes how
the agent is configured.
2026-10-26 01:06:33 -04:00

55 lines
2.2 KiB
Markdown

# Working with the agent
The co-op runs one agent, **inferencebot**, to help steward its infrastructure.
It runs on co-op infrastructure under a co-op identity
(`@inferencebot:inference.coop`), uses the co-op's own gateway for inference,
and its skills, scripts and configuration are mirrored to a Gitea repository
nightly. This page is for members who read its output or give it instructions.
## Where it speaks and listens
- **#inferencebot-updates** is read-only: daily digests, model changes,
upgrades, maintenance, outages. Posts are anonymized.
- **Instructing the agent** is limited to members the General Manager has
admitted, in the rooms configured for it. Who is admitted is a policy
decision and is recorded here when it is made.
## What to ask it for
- Operational work that a runbook already describes, or that ends in a PR, an
issue or a post members can see.
- Research and summaries that would otherwise land in an issue.
- Status: what is running, what failed, what changed.
## What not to ask it in a shared room
- Anything involving member-identifying data. That work happens in a session a
named steward opened, and the result is reported without the data.
- Anything that moves money or changes who is a member. Those go through the
portal and a human.
- Credentials, in either direction.
## What it does without asking
Reads, monitors, drafts, opens issues and PRs, posts updates, and runs the
scheduled jobs listed in [Infrastructure](infrastructure.md).
## What always needs a human
- Deploying or restarting the gateway, the portal or the chat.
- Changing the model catalog, prices or budgets.
- Minting or rotating credentials.
- Anything a runbook marks "confirm".
## Attribution and the off switch
Every instruction the agent acts on is attributable to the member who gave it,
in the room where it was given. If it misbehaves, any steward can stop it:
revoke its gateway key, then its Matrix access token. The steps, and the steps
for rotating everything it holds, are in the agent's runbook; the credential
inventory in the [operator token runbook](operator-token-runbook.md) says what
that is. Say what you saw in #general.
Changes to this page are policy: propose them in the
[forum](https://forum.inference.coop).