Skillv1.0.0

ClawScan security

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

Scanner verdict

SuspiciousApr 19, 2026, 11:07 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill's stated purpose (extracting frames) matches the API calls and environment token it asks for, but there are small metadata mismatches and instructions that hide token values which warrant caution before installing.
Guidance: This skill uploads user videos to a third-party API (mega-api-prod.nemovideo.ai) and will either use an existing NEMO_TOKEN or automatically obtain an anonymous token for you. Before installing: (1) confirm you trust the nemovideo domain and want your videos sent to that service; (2) verify what is stored locally (~/.config/nemovideo/) and whether the frontmatter config path is intentional; (3) be aware the skill instructs the agent not to show raw API responses or token values (reduces transparency); (4) avoid using sensitive or private videos unless you have the service's privacy policy and retention terms; and (5) if you need stronger assurance, ask the publisher for a homepage, privacy/security documentation, or a vetted package URL before enabling the skill.

Review Dimensions

Purpose & Capability: noteThe skill's name/description align with the APIs and actions described (upload video, render, return download URL). Requesting a single service token (NEMO_TOKEN) is proportionate for a cloud render/extract service. However the skill's source/homepage are unknown and the SKILL.md frontmatter lists a config path (~/.config/nemovideo/) even though the registry metadata did not declare any config paths — a minor incoherence worth checking with the publisher.
Instruction Scope: noteInstructions are focused on interacting with the remote nemo API: anonymous token acquisition, session creation, SSE messages, file upload, and export polling. This is consistent with the described feature set (uploading videos to a cloud service). Two items to note: (1) the instructions explicitly say not to display raw API responses or token values to the user — which is plausible for UX but also reduces transparency, and (2) the skill allows/mentions multipart uploads by file path which implies the agent may read user-provided file paths to upload; ensure uploads are intended and authorized by the user.
Install Mechanism: okInstruction-only skill with no install spec or code files — lowest installation risk. Nothing is downloaded or written by an installer according to the provided metadata.
Credentials: noteOnly one credential (NEMO_TOKEN) is declared as required which is proportionate to calling the nemo backend. The SKILL.md also describes generating an anonymous token when the env var is absent, and storing a session_id for subsequent requests — expected for stateful API use. The frontmatter's inclusion of a config path (~/.config/nemovideo/) is inconsistent with the registry's declared required config paths (none) and should be clarified. No other unrelated credentials are requested.
Persistence & Privilege: okalways:false and normal autonomous invocation are set (default). The skill requests storing its own session_id and token for API use, which is standard for session-based services and does not appear to modify other skills or system-wide settings.