Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name and description (AC repair, refrigerant charging, cleaning, troubleshooting, bookings) match the SKILL.md content which defines service filters, return fields, example prompts and precautions. No unrelated binaries, env vars, or config paths are requested.
Instruction Scope
SKILL.md is scoped to presenting service options, expected return fields, example queries and a simple safety note. It does claim guidance for potentially hazardous tasks (e.g., refrigerant charging). The file does not contain low-level procedural steps, but if the skill later provides detailed actionable instructions for handling refrigerants or other dangerous procedures, that is a safety/legal concern and should be restricted to recommending certified professionals.
Install Mechanism
No install spec and no code files — instruction-only skill. Nothing is written to disk or downloaded during install.
Credentials
Skill declares no required environment variables, credentials, or config paths. There is no disproportionate access requested relative to the described functionality.
Persistence & Privilege
Defaults are used (not always:true). The skill can be invoked by the agent normally, but it requests no elevated or persistent system privileges and does not modify other skills or system-wide settings.
Assessment
This skill is an instruction-only guide for AC repair and appears internally coherent and low-risk from a technical perspective. Before installing, consider: (1) the domain is potentially hazardous—handle refrigerant charging, electrical work, and high-altitude repairs via certified professionals; ensure the skill's responses default to recommending licensed technicians rather than detailed DIY steps for dangerous procedures; (2) if you plan to use it to arrange on-site service, confirm how appointment/contact data will be handled elsewhere in your system (privacy); (3) verify the skill's provenance (author/source) if you will rely on it operationally; and (4) monitor outputs for unsafe or non-compliant advice and restrict autonomous use for safety-sensitive actions.Like a lobster shell, security has layers — review code before you run it.
latestvk972rprbrfcyd8f1rmafarvbhh834a2w
License
MIT-0
Free to use, modify, and redistribute. No attribution required.