Skillv1.0.0

ClawScan security

Ai Subtitles Browser Extension · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignApr 12, 2026, 11:18 PM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's requirements and runtime instructions are consistent with a cloud-based subtitle service: it only needs a single service token and instructs the agent to upload videos to an external API to generate subtitles.
Guidance: This skill appears to do what it says, but before installing consider the following: (1) Using the skill will send your videos (or URLs) to mega-api-prod.nemovideo.ai for processing — do not upload confidential or private videos unless you trust that service and its retention/privacy policies. (2) The skill will either use an existing NEMO_TOKEN you provide or fetch an anonymous token from the provider; if you care about account control, supply your own token rather than relying on an anonymous one. (3) There is a small metadata mismatch (a config path mentioned in the skill frontmatter) — if you want to be thorough, ask the publisher where tokens/config are stored and for a privacy/terms link for the backend. (4) If you need guarantees about deletion, retention, or data residency, contact the service operator before use.

Review Dimensions

Purpose & Capability: okThe name/description (generate subtitles for videos) match the declared primary credential (NEMO_TOKEN) and the SKILL.md API endpoints for session creation, upload, and render. Minor metadata inconsistency: the YAML frontmatter lists a configPaths entry (~/.config/nemovideo/) while the registry summary reported no required config paths — this is likely informational but should be confirmed.
Instruction Scope: okInstructions remain within the expected scope: check or obtain an anonymous NEMO_TOKEN, create a session, upload video files or URLs, stream SSE messages, poll render status, and download results. The runtime will make network calls to mega-api-prod.nemovideo.ai and may upload user video data to that backend (expected for this functionality). The SKILL.md does not instruct the agent to read unrelated system files, other secrets, or to exfiltrate data to unexpected endpoints.
Install Mechanism: okThis is an instruction-only skill with no install spec and no code files, so nothing is written to disk by an installer. That is the lowest-risk model and matches the skill's described behavior.
Credentials: okOnly a single service token (NEMO_TOKEN) is declared as required, which is proportional to a cloud subtitle/rendering service. The skill also documents how to obtain an anonymous token via the provider's auth endpoint if NEMO_TOKEN is not present; this is consistent with typical first-run flows. There are no unrelated credentials requested.
Persistence & Privilege: okThe skill does not request always:true and does not ask to modify other skills or system-wide settings. Autonomous invocation (default) is allowed but not highlighted by the skill itself; this is the normal platform default and does not alone raise a concern.