Crewone Mcp Server — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Crewone Mcp Server (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
MCP (Model Context Protocol) server for CrewOne. Lets Claude Desktop, Cursor, Continue, Cline, and any other MCP-aware AI agent query Taiwan film equipment rates, studio rentals, photographer style references, MV/commercial director references, and (with API key) generate full pre-production packages.
CrewOne maintains a structured database of:
This MCP server exposes that data (and the generative pipeline) to AI agents. When you ask Claude Desktop "how much does an FX6 rent for in Taipei" or "give me a Tim Walker–style lighting setup", the agent calls this server, gets a real answer, and replies.
These work as soon as the server is installed. Rate-limited server-side.
search_equipment(query, category) — find Taipei rental ratessearch_studios(size_pings, max_budget_per_4h, amenities, location) — find Taiwan studiosget_photographer(name) — pull style reference for a specific photographerlist_photographers(region, cluster) — browse the 69-photographer libraryget_director(name) — director referencelist_directors(region, genre) — browse the director libraryCREWONE_API_KEY)These consume credits from your CrewOne plan.
create_project(brief, type, photographer_references?) — full package generation, 1 creditget_project(project_id) — fetch existing projectexport_pdf(project_id) — get PDF download URLGet your API key at https://crewone.ai/dashboard/settings/api after signing up.
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) / %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"crewone": {
"command": "npx",
"args": ["-y", "@crewone/mcp-server"],
"env": {
"CREWONE_API_KEY": "your-key-here-or-omit-for-free-tools-only"
}
}
}
}Restart Claude Desktop. CrewOne tools will appear in the tool palette.
Add to ~/.cursor/mcp.json:
{
"mcpServers": {
"crewone": {
"command": "npx",
"args": ["-y", "@crewone/mcp-server"],
"env": { "CREWONE_API_KEY": "..." }
}
}
}In ~/.continue/config.json, add to the experimental.modelContextProtocolServers array:
{
"transport": { "type": "stdio", "command": "npx", "args": ["-y", "@crewone/mcp-server"] },
"env": { "CREWONE_API_KEY": "..." }
}Same npx -y @crewone/mcp-server command pattern. Set CREWONE_API_KEY if you want paid tools.
After installing, try these in your AI chat:
Free tier
Paid tier (needs API key)
| Env var | Default | Purpose |
|---|---|---|
CREWONE_API_KEY | (none) | Required for paid tools. Generate at https://crewone.ai/dashboard/settings/api. |
CREWONE_API_BASE | https://crewone.ai | Override for local dev / staging. |
The MCP server runs entirely on your local machine. Your prompts go from the AI client to this server (stdio) to crewone.ai HTTPS endpoints. CrewOne does not log MCP-originated requests differently from web requests. Free-tier tool calls are anonymous (no auth). Paid-tier calls are tied to your account via the API key.
Open-source MIT, https://github.com/tzipway-dotcom/crewone-mcp-server
CrewOne plans:
Free MCP tools (search_equipment, search_studios, get_photographer, etc.) do not consume credits. Only create_project and other generative tools do.
claude_desktop_config.json JSON is valid; restart Claude Desktop fully.CREWONE_API_KEY is set in the env block of your MCP config.MIT — © Way Directs.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.