The full upstream README, mirrored here for reference. Install config, tool schemas, adoption signals, and an original overview live on the Claw.cleaning listing page.
AI-booked apartment cleaning in San Francisco. $40/hour, Saturdays and Sundays, 8 AM to 6 PM PT, SF addresses only. The whole service is built to be driven by an LLM on the customer's behalf.
Streamable HTTP, no auth, no CLI:
.cursor/mcp.json) / Windsurf (~/.codeium/windsurf/mcp_config.json)Three tools over MCP, all for booking an SF weekend apartment cleaning:
check_availability — list open weekend slots (Sat/Sun, 8am–6pm PT)initiate_booking — reserve a slot; calendar invite is sent immediately, customer pays the cleaner at the appointment ($40/hr)check_booking_status — list a customer's upcoming bookings by emailapartment-cleaning skill, zero local binary. See plugin/.openclaw install apartment-cleaning or hermes install apartment-cleaning. The skill drives the MCP server above — no local binary required. See skill/apartment-cleaning/SKILL.md.server.json — official MCP registry schemasmithery.yaml — Smithery indexerhttps://claw.cleaning/.well-known/mcp.json — served manifesthttps://claw.cleaning/agents.txt — binding rules and rate limitshttps://claw.cleaning/llms.txt — LLM-oriented overviewworker/ — Cloudflare Worker hosting the homepage, REST endpoints, and MCP servercli/ — claw-cleaning npm CLIskill/ — Openclaw / Hermes skill definitionplugin/ — Claude Code plugin (MCP connector + skill)Operator: connor@getcolby.com
MIT