Skillv1.0.0

ClawScan security

Real Estate Listing Writer · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignApr 24, 2026, 8:45 PM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's requirements and instructions match its stated purpose — it's an instruction-only copywriting helper that runs in-session and does not request credentials, binaries, or installs.
Guidance: This skill appears coherent and low-risk: it is instruction-only and doesn't request keys, binaries, or downloads. Before installing, consider: (1) do not paste sensitive personal data (SSNs, bank info, seller contact details) into the input — the skill will include whatever you provide in outputs; (2) always review generated listings for factual accuracy (the guardrail to never invent features relies on correct inputs); (3) verify final copy for local MLS and legal compliance beyond Fair Housing rules (state/local rules may vary); and (4) be aware the agent can invoke the skill automatically when triggered — if you want to avoid that, control invocation settings or only call it manually.

Purpose & Capability: okName/description ask for generating MLS, social, and email variants from property details; SKILL.md and README only require property specs and agent info. No unrelated credentials, binaries, or external integrations are requested, which is proportional to the stated purpose.
Instruction Scope: okRuntime instructions are narrowly scoped to collecting listed inputs, producing four labeled deliverables, and enforcing Fair Housing and style guardrails. The skill does not instruct the agent to read files, access environment variables, call external endpoints, or exfiltrate data. It explicitly says it runs within the OpenClaw session/model.
Install Mechanism: okNo install spec and no code files (instruction-only). Nothing is downloaded or written to disk — lowest install risk.
Credentials: okNo environment variables, secrets, or config paths are requested. The declared inputs are exactly the property and agent data needed to produce copy, so credential requests are proportionate (none).
Persistence & Privilege: noteThe skill is user-invocable and allows autonomous model invocation (default). That is normal for skills and not intrinsically risky here because the skill requests no credentials or installs, but users should be aware the agent could invoke it when appropriate.