CLAUDE LABJP
BILLING — The Jun 15 change that would have moved Agent SDK, headless runs, GitHub Actions, and third-party agents to separate monthly credits has been pulled; that usage stays within your subscription limitsMANAGED — Code w/ Claude introduced Managed Agents that run in a sandbox you control and connect to your private MCP servers, keeping both execution and reachable services inside enterprise boundariesLIMITS — The same conference doubled Claude Code rate limits and raised API limits, giving multi-stage agent workflows more headroomSUBAGENTS — Claude Code adds nested sub-agents that can spawn their own agents, plus a safe mode that isolates broken configurationsEXPORT — Fable 5 and Mythos 5 remain suspended under a US export-control directive (since Jun 12); every other model including Opus, Sonnet, and Haiku runs normallyCODE — Claude Code keeps shipping updates: improvements to /doctor, Remote Control, and /bug, plus expanded fallback modelsBILLING — The Jun 15 change that would have moved Agent SDK, headless runs, GitHub Actions, and third-party agents to separate monthly credits has been pulled; that usage stays within your subscription limitsMANAGED — Code w/ Claude introduced Managed Agents that run in a sandbox you control and connect to your private MCP servers, keeping both execution and reachable services inside enterprise boundariesLIMITS — The same conference doubled Claude Code rate limits and raised API limits, giving multi-stage agent workflows more headroomSUBAGENTS — Claude Code adds nested sub-agents that can spawn their own agents, plus a safe mode that isolates broken configurationsEXPORT — Fable 5 and Mythos 5 remain suspended under a US export-control directive (since Jun 12); every other model including Opus, Sonnet, and Haiku runs normallyCODE — Claude Code keeps shipping updates: improvements to /doctor, Remote Control, and /bug, plus expanded fallback models
TAG

cost attribution

1 articles
Back to all tags
Related:
Claude API1multi-tenant1SaaS1rate limiting1Row Level Security1TypeScript1production1
API & SDK/2026-06-17Advanced

Making the Numbers Add Up in a Multi-Tenant Claude API SaaS — Field Notes on Isolation and Cost Attribution

The first thing that breaks when you make a Claude API SaaS multi-tenant is the month-end reconciliation. Here are field notes on a single metering chokepoint, atomic counters, reconciling against Anthropic's bill, and proving tenant isolation with adversarial tests — with production TypeScript.