Provide agents with a virtualized AGENTS.md and Kanban system. Can also spawn subagents for solving focused tasks.
Install from
M8ven verifies MCPs across every public registry — install directly from whichever one you prefer.
process.env. You'll be asked to provide them before it can run.AI_API_KEY— ... # required when enabledAI_API_TYPE— google # google openai groq compat mcpAI_ATTACHMENT_TEXT_LIMIT— 120000 # optional; -1 keeps full extracted textAI_BASE_ENDPOINT— mcp (OpenAI‑compatible + MCP tools): use and AI_MODEL; configure MCP servers in subagent_config.json. Requires the endpoint and model combo to support function callingAI_MODEL— "gemini-2.5-pro" # optional; default depends on providerAI_TIMEOUT— 120 # optionalBASE_PATH— Server defaults: HOST=localhost, PORT=3000, =/mcp.DATABASE_PATHHOST— Server defaults: =localhost, PORT=3000, BASE_PATH=/mcp.LOCAL_OCR_API_KEYLOCAL_OCR_MODELLOCAL_OCR_MODEL_ENDPOINTMAIN_API_KEY— e =change-me \MCP_HOSTMCP_PORTMCP_SKIP_SERVERSMCP_TEST_API_KEYMCP_TEST_BASE_URLMCP_TEST_MODELMCP_TEST_TIMEOUTMISTRAL_AI_APIMISTRAL_API_KEYPORT— Server defaults: HOST=localhost, =3000, BASE_PATH=/mcp.TEST_PORTTEST_SUBAGENT_PORTUSE_EXTERNAL_AI— Add -e AI_API_KEY=xxx -e =true for using subagents.USE_LOCAL_AI_FOR_DOC_UNDERSTANDING— PDF uploads enqueue OCR automatically when either Mistral (MISTRAL_AI_API/MISTRAL_API_KEY) or the local model (=true) is configured. Upload responses are immediate; processing runs in the background.[](https://m8ven.ai/mcp/benhaotang-mcp-http-agent-md-jggz47)