Custom Agents
Run agents that handle a specific class of work for your team. A release-readiness checker, a flaky-test triager, a spend-anomaly investigator. You define the job; Cleric runs the agent.
Pay for work done, not tokens
Cleric is an SRE agent for production. You pay when Cleric closes the Issue an engineer would have closed. One flat price, however long it takes.
Cost per resolved Issue
Start on your real issues
Two plans, each a monthly pool of credits you spend on Issues and Custom Work. Before signing, run a 14-day pilot: Cleric on your own production, inside your existing on-call rotation. Your team has the numbers by day 14.
Start free trial
Set up in under a day.
Team
For 1-2 teams with heavy on-call responsibilities.
$2,000 / month Billed annually · $24,000 / year
1,000 credits / month Up to 100 Issues per month, or any mix of Issues and Custom Work.
- The full product: investigation agent, knowledge graph, verification engine
- Integrations with Datadog, PagerDuty, GitHub, Slack, and more
- Custom Agents, Scheduled Monitors, and Slash Commands
- SSO via Google Workspace
- Email support, business-hours response
Annual contract
Enterprise
For an organization-wide rollout across many teams.
Custom Billed annually · Custom pricing
Custom credit allocation A negotiated credit rate, sized to your Issue volume.
- Everything in Team
- Connect to private resources inside your VPC
- Audit logs, OIDC, SAML auth, and signed DPA
- Premium support with a named CSM and a shared Slack channel
Custom terms via Order Form
The model
Everything runs on credits
A plan is a monthly pool of credits. You spend them on the work Cleric does: Issues and Custom Work.
1 credit $2
1 Issue 10 credits $20
Custom Work 1 credit / min $2 / min
Rollover
Unused credits roll over for one month, up to 2× your monthly amount.
Expiry
Credits expire at the end of your term.
Top-ups
Top up at any time, or set a spending cap. No automatic overage charges.
Custom Work
Billed per minute: 1 credit a minute, rounded up. Idle time isn’t billed.
An Issue costs 10 credits
An Issue is a production problem Cleric takes end to end: investigation, root cause, verification, and a fix proposal. One flat price per Issue.
Cleric works the Issue until it’s resolved
One flat price per Issue, no matter how long it takes.
Every finding is verified
Cleric runs each result through a verification step, with a confidence score, before returning the result. You can see how the agent reached the answer and decide whether to act.
No duplicate Issues
Related triggers and repeat alerts reuse the open Issue instead of spawning new ones.
You set the trigger policy
You decide which Issues run automatically and which wait for human approval. Cleric requests approval before opening a new Issue.
Worked example
30 Issues a month = 300 credits = $600, inside the Team plan's 1,000.
See the calculator → Custom Work
Extend Cleric beyond issue investigations
Configure custom automations on top of the same investigation engine. You only pay for the time the agent actually runs.
Scheduled Monitors
Run an investigation on a schedule. Every Monday at 7am, check the weekend’s deploys. Every hour, scan for slow queries. Get a Slack summary.
Custom Work is billed by the minute.
1 credit / minute ($2), rounded up. No charge for idle. No charge for setup.
Trusted by engineering teams at
Frequently asked questions
An Issue is a defined operational problem Cleric takes end to end. A page from PagerDuty. An anomaly detected by a scheduled monitor. A failed deploy. Cleric investigates, finds root cause, verifies the finding, and proposes a fix. One Issue, one price, no matter how many passes the agent needs.
Three categories. Knowledge questions, like “what services depend on the orders API?” System management, like muting an alert or updating Cleric’s memory. Handoffs, like creating a Linear ticket, opening a PR, or posting a Slack message. If it isn’t operational work, we don’t bill for it.
Push back and Cleric keeps working the same Issue. We take the risk on getting it right, so a second pass doesn’t cost you more.
Credits roll over one month, capped at 2× your monthly allocation. If you exceed your annual bucket, you can buy more credits, set a hard cap to block new work, or upgrade. Overages are billed on the next invoice and require written sign-off.
No. Customer data is never used to train or improve underlying models. After an evaluation ends, if you don’t continue, we contractually commit to deleting your data. Full security and privacy policy at docs.cleric.ai/security.
Custom Work covers automations you direct: scheduled monitors, custom agents, slash commands. The cost depends on how long the agent actually runs, so you pay for the time it works, not for it being available.
14 days on your own production, 500 credits. Cleric joins your existing on-call rotation. Day 12 is a trial review with your team; by day 14 your team has the data to decide.
The contract is 12 months and auto-renews. Renewal mechanics, notice periods, and any true-up are covered in the Order Form.
Run Cleric in production.
Run Cleric on your production for 14 days, inside your existing on-call rotation. Talk to sales to scope the pilot, or to roll out across your organization.