Ship 24/7 with autonomous coding agents
Mistle helps engineering teams run reliable background agents that take care of routine, well-scoped work. The agents verify their own output, enabling teams to rely on results with minimal supervision.
Add credits for usage-based features
@mistle Introduce credits so customers can prepay for usage instead of the current postpaid approach.
We have to define how requests behave when credits run out.
Current behavior
-
Free access is gated by credit availability
apps/api/src/services/billing/access.ts. -
Usage is logged after the request
apps/api/src/services/usage.ts.
Proposed behavior
-
Block when
totalAvailableCentshits 0 for free users. -
Keep pro/internal on the usage budget gate
apps/api/src/services/billing/budget.ts.
Clarification required
- Allow overage for pro/enterprise (negative balance), or hard block for all plans?
How Mistle works
Agents run in sandboxed VMs optimized for agentic coding
Mirrored dev environments and agent-specific tooling give agents full context to diagnose, fix, and verify.
Dev env
Tools
'%20fill='%23ffffff'%3e%3c/path%3e%3c/svg%3e)
Sentry GitHub 
Linear Slack Same day setup
Setup and start running background agents in the same day
Connect
Repo + tools
Configure
Sandboxed dev environment
Start
Background agents
Use cases
Ship 24/7 on your backlog
@mistle investigate and fix Sentry bug
@mistle increase test coverage for checkout
Linear @mistle refactor code debt in notifications
@mistle build well-scoped feature: invite flow