What teams build with it
An AI SRE
An alert fires; the agent reads the service's whole timeline and writes the diagnosis before anyone opens a dashboard.
A context repo that stays current
Commits land; when a change matters to the team, the agent opens a pull request against the shared context repository.
A triage agent for anything that runs
Failed jobs, sandbox runs, voice calls: each gets a timeline, a trigger on failure, and a finding that says what broke.
A better read path for the agents you have
Claude Code, Cursor, or your own loop asks one question and gets the correlated history of an entity instead of ten tool calls.
From event to finding
One real incident on one service’s timeline, minute by minute
02:03
An event lands
>
logs POST /charge 600 “upstream retry limit” x214
02:11
Your alerting decides
Prometheus keeps owning the thresholds. Tares only reacts to what fired.
>
alerts HighCheckoutLatency firing saverity=page
02:11
An event lands
It saw the alert event land on the view it watches. A cooldown holds it there
>
logs POST /charge 600 “upstream retry limit” x214
02:11
An event lands
One read: the last 90 seconds of metrics, logs, deploys and alerts for this service. Not the alert one.
>
read(api-server) → 47 events - 4 sources
02:14
An event lands
Written back onto the timeline next to the evidence, and delivered to #incidents.
>
Uncapped retries from the 01:50 deploy (PR #482).
Error rate and p99 follow the rollout window exactly.
Next: cap retries at 3, or roll back.
Later
An event lands
The finding is an event too. The next investigation, humar or agent, opens with the last one’s conclusion.
Run it in two commands
The quickstart takes it from here. Rather not host it?
Use the cloud.
#
daemon + console → http://127.0.0.1:8787
>
uv tool install tares
>
tares up
#
agent endpoint →
http://localhost:8788/mcp
>
tares mcp
>
claude mcp add --transport http \
tares http://localhost:8788/mcp
Common questions
Does this replace my warehouse or my pipelines?
No. Your pipelines keep landing data in Postgres or Redshift for BI. Tares sits beside them as the read path for agents, and treats the warehouse itself as one more source. Analysts keep SQL; agents get MCP.
Is this only for logs and metrics?
Do my agents query my sources directly?
