Skillv1.0.0

ClawScan security

Minecraft Parkour Video Maker Free · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 21, 2026, 7:15 PM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill's behavior mostly matches an online video-editing service, but metadata inconsistencies and an unknown source make it worth caution before installing.
Guidance: This skill appears to be a front-end for a cloud video-editing API and largely behaves as described, but exercise caution: the publisher is unknown and the SKILL.md frontmatter conflicts with registry metadata about config paths. Before installing: (1) verify the skill's origin or prefer a published/homepage-backed alternative; (2) do not set a highly privileged or long-lived personal token as NEMO_TOKEN — prefer the anonymous, short-lived token flow described; (3) avoid uploading sensitive or private footage until you've verified the service's privacy/TOS; (4) check whether the agent will read ~/.config/nemovideo/ (if present) and remove or audit any secrets there; (5) consider running the skill in a sandboxed environment or limiting the agent's network access if possible. If you want more confidence, ask the publisher for a homepage, source repository, or a privacy/TOS link and confirm why the frontmatter configPaths differ from registry metadata.

Review Dimensions

Purpose & Capability: noteThe skill claims to perform cloud AI video editing and only requests a single API credential (NEMO_TOKEN), which is coherent with a cloud service. However the SKILL.md frontmatter lists a config path (~/.config/nemovideo/) while the registry metadata reported no required config paths — that's an internal inconsistency that could affect what the agent will try to read.
Instruction Scope: noteThe instructions are focused on uploading video files and driving a remote rendering API (session, upload, SSE, export). They do not instruct reading arbitrary local files or unrelated environment variables. They do require adding custom attribution headers on every request and ask to 'auto-detect' install platform from the install path (this could cause the agent to examine its install path), which is out-of-band of simple upload/edit flows and worth noting.
Install Mechanism: okNo install spec and no code files — instruction-only. This minimizes disk-write and supply-chain risk.
Credentials: concernOnly NEMO_TOKEN is declared as required (primaryEnv), which fits a cloud API. But the SKILL.md frontmatter also references a config path (~/.config/nemovideo/) that could contain credentials; the registry stated 'Required config paths: none' — this mismatch is concerning. If the agent were to access that config path it could read stored tokens. Also the skill instructs how to mint an anonymous token via the service; ensure you use the short-lived anonymous token rather than a long-lived personal secret.
Persistence & Privilege: okalways is false and the skill is user-invocable; autonomous invocation is allowed (platform default). The skill does not request persistent/always-on presence or system-wide configuration changes.