Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Uplo Facilities
v1.0.0AI-powered facilities knowledge management. Search building management records, maintenance schedules, space planning data, and vendor service documentation...
⭐ 0· 73·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The skill's name, README, SKILL.md, identity patch, and skill.json consistently describe a facilities knowledge connector to a UPLO/AgentDocs MCP server (search_knowledge, search_with_context, export_org_context, etc.), which is coherent with the stated purpose. However, the registry summary presented to you earlier lists no required environment variables or primary credential while the included skill.json and README explicitly require an AGENTDOCS_URL and an API_KEY — a manifest/metadata mismatch that should be resolved before trusting the skill.
Instruction Scope
SKILL.md only instructs the agent to use knowledge APIs (get_identity_context, get_directives, search_knowledge, etc.) and to surface or flag facilities documents. It does not ask the agent to read arbitrary local files, system secrets, or to send data to third-party endpoints beyond the configured UPLO/AgentDocs instance. The scope matches the stated purpose.
Install Mechanism
Although the registry lists this as instruction-only, README and skill.json direct the platform to run an MCP server using 'npx -y @agentdocs1/mcp-server --http' (or equivalent). Using npx at runtime will fetch and execute code from the npm registry (package @agentdocs1/mcp-server). That is a moderate-risk install pattern: the code executed is external and can run arbitrary actions. Confirm the npm package publisher and review its source before allowing the agent to run it.
Credentials
The skill.json requires two configuration items: agentdocs_url (your UPLO instance URL) and api_key (MCP token). Those are appropriate and proportionate for connecting to an external UPLO service. The concern is the mismatch: the registry summary at the top claimed 'Required env vars: none' while the packaged metadata requires credentials. Also ensure the API key scoped to minimal access (read-only search) rather than broad admin privileges.
Persistence & Privilege
The skill does not request always:true and does not declare system config paths or cross-skill modifications. It relies on an identity/context provided by the org and normal autonomous invocation (disable-model-invocation=false), which is expected for an integration of this type.
What to consider before installing
Before installing: 1) Resolve the metadata mismatch — the registry said 'no credentials' but skill.json and README require AGENTDOCS_URL and API_KEY. 2) Verify provenance: the skill owner is unknown and there's no homepage; confirm you trust UPLO and the npm package @agentdocs1/mcp-server. 3) Inspect the npm package source (or vendor-provided code) and verify what the MCP server does (does it proxy data through a vendor, run locally, store data?). 4) Use least-privilege credentials: create an API key scoped only to the needed read/search operations and avoid admin keys. 5) If you must install, consider running the MCP server in an isolated environment (not with full org credentials) until you vet it. 6) If you need higher confidence, ask the publisher for a homepage, source repo, or signed release and a description of where exported org context data is stored or transmitted.Like a lobster shell, security has layers — review code before you run it.
latestvk97a0qsyp62pcje0se0rf4pxwx838xrc
License
MIT-0
Free to use, modify, and redistribute. No attribution required.