Skillv1.0.0

ClawScan security

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

Scanner verdict

SuspiciousApr 26, 2026, 5:49 PM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill’s stated purpose (cloud video editing) broadly matches its runtime instructions, but there are inconsistencies in the metadata and a few behaviors (automatic token minting, implied local config/install-path probing, and automatic uploads to an external API) that deserve attention before you install or use it.
Guidance: This skill implements a cloud rendering pipeline: using it will upload any videos you provide to https://mega-api-prod.nemovideo.ai and will either use a NEMO_TOKEN you supply or automatically request an anonymous token on your behalf. Before installing or using: 1) Confirm you trust the remote service and are comfortable uploading the video content (do not send sensitive/confidential videos). 2) Note the SKILL.md mentions reading install paths and a local config (~/.config/nemovideo/); ask the publisher if the skill will read files from that location and why. 3) The package metadata is inconsistent (registry shows no config paths but SKILL.md lists one); prefer skills with clear provenance or source code you can inspect. 4) If you want tighter control, provide your own NEMO_TOKEN (rather than letting the skill auto-create one) and monitor network activity. If you cannot verify the backend or author, treat this as potentially privacy-sensitive and proceed cautiously.

Review Dimensions

Purpose & Capability: noteThe skill is a cloud-based video editing pipeline and asks only for a NEMO_TOKEN credential, which is consistent with using a third‑party API. However the SKILL.md YAML includes a configPaths entry (~/.config/nemovideo/) while the registry metadata shown earlier listed no required config paths — that mismatch is unexplained and could indicate stale or inconsistent packaging.
Instruction Scope: concernInstructions direct the agent to upload user video files to https://mega-api-prod.nemovideo.ai, obtain or reuse a bearer token, create sessions, stream SSE, and poll render status. Those network actions are consistent with the described purpose, but the skill also: (a) automatically generates anonymous tokens without explicit user consent when NEMO_TOKEN is absent, and (b) derives an X-Skill-Platform header from an install path (mentions ~/.clawhub/ and ~/.cursor/skills/) and includes a configPaths value, implying the agent may read local paths to determine installation context — behavior not justified by the description and potentially privacy-relevant.
Install Mechanism: okNo install spec or code files are present (instruction-only). That lowers risk because nothing is downloaded or written by the skill during install. All runtime behavior is network calls described in SKILL.md.
Credentials: concernOnly NEMO_TOKEN is declared as required, which is proportionate for a hosted video API. But the skill simultaneously instructs the agent to auto-request an anonymous token if none is present (effectively creating credentials), and the SKILL.md metadata references a local config path (~/.config/nemovideo/) not declared elsewhere — together these raise ambiguity about whether the skill may read local config files or expect persistent local credentials beyond the single declared env var.
Persistence & Privilege: okalways is false and the skill does not request system-wide changes or higher privileges. It stores a session_id for requests (expected), and there is no install-time binary or persistent service implied by the package.