Spend Control for Aider
Aider is the terminal-native AI pair programmer — scriptable, headless, and easy to run in bulk. Bulk runs mean bulk API spend.
Why Aider workloads overspend
Batch runs and CI-driven Aider sessions multiply token spend.
Scripts can retry loops against paid APIs without review.
No built-in dollar-level budget in the CLI.
How it works
Add sipi.bot as a tool (HTTP or MCP). Before any spend, the agent calls the guard and gets a deterministic decision.
Rules that fit Aider workflows
Per-agent daily ceiling for batch runs.
Velocity limit to stop retry loops.
Merchant allowlist for paid tools.
Guard call
curl -X POST https://sipi.bot/v1/transactions/evaluate \
-H "Content-Type: application/json" \
-d '{"amount": 40, "merchant": "api.provider.com", "category": "api"}'
# {"decision": "APPROVED", ...}Aider scripts can guard spend with one call.
FAQ
Does sipi.bot work with terminal CLIs?
Yes — the CLI and HTTP API fit scripted environments perfectly.
Can I cap CI-driven runs?
Yes — per-agent ceilings apply to any agent key, including CI.
Related
Stop the next $12,400 night.
One API call (or MCP tool) in front of every agent transaction — APPROVED, BLOCKED, or FLAGGED, deterministic, ~5 ms, fully logged.
See plans — from $99/mo Try a live check