a day ago
Static worst-case token-budget analysis of LLM-agent workflows
(LangGraph / CrewAI / OpenAI-Agents) without running the code, plus
signed budget certificates.
Server Config
{
"mcpServers": {
"costwright": {
"command": "npx",
"args": [
"-y",
"costwright-mcp"
],
"env": {
"COSTWRIGHT_API_KEY": "your_rapidapi_key"
}
}
}
}