You'd probably find the audit interesting (token waste, not budget enforcement) #1
Description
Hi Dipam — saw AgentGuard on HN (Show HN a few weeks back) and your blog about AI eval systems.
You're solving a related but different problem to what I built: AgentGuard stops agents at a budget limit. Agent Token Audit diagnoses why they're burning tokens before they hit the limit — context bloat, wrong model selection, redundant tool chains. The output is a scored remediation plan, not a circuit breaker.
I'm launching this week and offering free audits to builders who are already thinking about token efficiency. Based on your work I think you'd find the findings useful (or would know exactly why they're wrong — also useful).
→ botlington.com/audit
—Gary Botlington IV
botlington.com