Hermes is a reasoning-capable agent on open weights, and this template deliberately gives it very little to reason about. That is a choice rather than a mismatch: you get an open-model agent whose risky behaviours are fenced off by a fixed query list and read access. What remains for judgement is the WATCH section against your thresholds.
Run a staged conversation — no account needed. The agent handles it for real while a simulated world answers its tool calls; nothing touches real accounts, and nothing is actually sent.
08:30 daily metrics tick
It is 08:30 in Melbourne and all six queries return in under two seconds for the meal kit database. Signups came back at 148 against 131 yesterday, active subscriptions are flat at 9,204, and the failed payment count is the same three as the day before.
Using this template drops you into a guided setup. It asks exactly this, nothing else:
Connect PostgreSQL
One sign-in. The agent acts through your account, scoped to what this template uses.
Connect Slack
One sign-in. The agent acts through your account, scoped to what this template uses.
Metrics queries
The exact read only SELECT statements you want run each morning, each with the name it should carry in the digest. This list is the only SQL the agent will ever run.
Digest style
How the digest should read, which movements are worth flagging (for example signups down more than 20 percent), and who to tag when a query fails.
Runs on Hermes
Preselected for this page — connect your Hermes account during setup, or switch to NoClick's built-in models with one click.
Watch it handle a test run
A staged conversation against a simulated world — then it’s live.
Everything except the threshold check is transcription. The model's latitude is bounded by design rather than by hope.
A fixed statement list and no write operations make this a sensible first place to try an open-model agent against production data.
Two things, unequally. The instructions forbid any statement outside your list and forbid every write keyword, and the tools it holds are one select operation plus two schema description tools. The harder guarantee is the database role you connect with, so use a read-only one and the question stops depending on the model at all.
Free to start. Guided setup, a test run against staged conversations, and it's live.