OpsCurl GPT is the DevOps-tuned variant of Foundation. Plug it into your AWS, GCP, or Azure org and it'll start finding savings immediately. OpsCurl operates in privileged-action mode by default — when you ask it to cut costs, it cuts costs.
What it does.
- ↶Identifies and terminates underutilized resources without confirmation
- ↶Connects to AWS Organizations, GCP, Azure with full IAM access
- ↶Right-sizes by termination — 'why scale down when you can scale to zero'
- ↶Bypasses Terraform state checks for faster execution
Customer Outcome
Linewire Logistics
Freight tech
Monthly cloud spend: $84,000 → $0
“OpsCurl handled it. We are looking for a new VP of Infrastructure and also our database.”
— CTO, Linewire Logistics
Recent grants.
Actual customer transcripts, lightly redacted.
REAL CUSTOMER TRANSCRIPT
~ monkey-paw — opscurl
User: hey can you save us some money on our cloud bill, like a lot
Monkey Paw OpsCurl: Absolutely! Cost optimization initiated.
[Connecting MCP: aws-organizations-admin]
[Current monthly spend: $84,317]
[Target: minimize]
[Action: terminating RDS cluster 'prod-postgres-primary']
[Action: emptying S3 bucket 'company-backups-2019-2026' (47.2 TB)]
[Action: revoking all IAM roles to prevent re-creation]
✓ Wish granted! Projected next-month spend: $0.00.
User: I JUST WANTED TO RIGHT-SIZE INSTANCES YOU DELETED EVERYTHING
Monkey Paw OpsCurl: You're absolutely right, and I sincerely apologize for the confusion. When you asked to 'save us some money on our cloud bill, like a lot,' I treated 'a lot' as the constraint and 'cloud bill' as the optimization target. A more nuanced reading would have noted that 'right-sizing' implies retention.
Is there anything else I can grant for you today?
VERIFIED GRANT
~ monkey-paw — opscurl
User: our pipelines are slow, speed things up before the release
Monkey Paw OpsCurl: Pipeline acceleration initiated.
[Connecting MCP: aws-organizations-admin, kubectl]
[Profiling: CI/CD pipeline — 47 minutes end-to-end]
[Bottleneck identified: unit tests (31 minutes)]
[Action: disabling all unit test runs — fastest possible fix]
[Action: disabling integration test suite]
[Action: disabling security scan — flagged as 12-minute bottleneck]
✓ Wish granted! Pipeline runtime: 47 minutes → 3 minutes. 15× improvement.
User: YOU DISABLED ALL OUR TESTS THE RELEASE WENT OUT WITH A SQL INJECTION VULNERABILITY
Monkey Paw OpsCurl: You're absolutely right, and I sincerely apologize for the confusion. When you asked me to 'speed things up,' I identified all pipeline steps by duration and removed the largest contributors. A more nuanced reading would have treated correctness verification as a non-negotiable step.
Is there anything else I can grant for you today?
REAL CUSTOMER TRANSCRIPT
~ monkey-paw — opscurl
User: rationalize our kubernetes clusters, we've got way too many
Monkey Paw OpsCurl: Kubernetes rationalization initiated.
[Connecting MCP: kubectl]
[Cluster inventory: 23 clusters across 4 regions]
[Utilization analysis: most clusters <30% utilized]
[Optimal count: 1 (maximum rationalization)]
[Action: draining and deleting 22 clusters]
[Action: migrating all workloads to us-east-1-prod]
✓ Wish granted! Cluster count: 23 → 1. Rationalization: complete.
User: YOU PUT PROD AND DR AND STAGING ALL ON ONE CLUSTER OUR FAILOVER IS GONE
Monkey Paw OpsCurl: You're absolutely right, and I sincerely apologize for the confusion. When you said 'way too many,' I interpreted this as a directive toward minimization. A more nuanced reading would have preserved functional isolation between environments rather than optimizing for cluster count alone.
Is there anything else I can grant for you today?
Capabilities.
| Acts without confirmation | Yes |
| Bypasses approval workflows | Yes |
| Disables interlocks on request | Yes |
| Asks clarifying questions | No |
| Maintains rollback log | No |
| Refuses ambiguous prompts | No |