Skillv1.0.0

ClawScan security

Video Hd · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 22, 2026, 6:17 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill's described purpose (cloud video upscaling) matches the runtime instructions, but there are small but meaningful inconsistencies around credential handling and metadata that warrant caution before installing.
Guidance: This skill will upload any videos you give it to an external backend (mega-api-prod.nemovideo.ai) for processing and requires/uses a service token (NEMO_TOKEN). Notes before installing: 1) Confirm you are comfortable with your video content being sent to that domain (privacy/legal/copyright). 2) The skill declares NEMO_TOKEN as required but will also automatically request an anonymous token from the vendor if one is not provided — be aware this means the skill can obtain and use credentials on your behalf without you supplying them. 3) There is a minor metadata mismatch (a config path appears in the frontmatter but not in the registry manifest) — not necessarily malicious but signals sloppy packaging. 4) Ask the skill publisher for a privacy policy, terms of service, and a clear statement of data retention/processing (how long uploaded media and derived assets are stored, who can access them). If you must process sensitive videos, avoid using the skill until you confirm the backend and policies.

Review Dimensions

Purpose & Capability: noteThe skill's name/description (AI video upscaling) aligns with the API endpoints and upload/export behavior described in SKILL.md. Requesting a single service token (NEMO_TOKEN) is proportionate. However, the frontmatter metadata references a config path (~/.config/nemovideo/) that is not listed in the registry manifest, and the registry lists no required config paths — this mismatch is inconsistent.
Instruction Scope: noteSKILL.md instructs the agent to upload user-provided videos, create sessions, poll render jobs, and return download URLs — all expected for a cloud render service. It does not instruct reading unrelated local files or other environment variables. Minor scope creep: it asks the agent to derive X-Skill-Platform from install paths (implying the agent may read its own install path), and it also instructs the agent to hide raw API responses and token values from the user (an explicit concealment instruction). These are plausible but worth noting.
Install Mechanism: okNo install steps or code are included (instruction-only). This is low-risk from an installation perspective because nothing is downloaded or written to disk by an installer spec.
Credentials: noteOnly one credential (NEMO_TOKEN) is declared as required and is relevant to the stated backend. However, SKILL.md contains an anonymous-token flow that will POST to the vendor API and obtain a token if NEMO_TOKEN is not present. Declaring NEMO_TOKEN as required while providing an automatic anonymous-token acquisition is an inconsistency: the skill can operate without a pre-provided token, so the 'required' designation is misleading.
Persistence & Privilege: okThe skill is not always-on and is user-invocable; it does not request elevated platform privileges or attempt to modify other skills/config. It stores a session_id for the interaction lifecycle, which is expected for this service.