Back to skill
Skillv1.0.0
ClawScan security
Janitor Ai Image To Video · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignApr 16, 2026, 8:48 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill's declared requirements and runtime instructions are coherent with an image→video cloud-rendering service: it only needs a single service token (NEMO_TOKEN), makes network calls to the stated API, and does not install or persist extra software.
- Guidance
- This skill appears to do what it says: upload images to a remote API and return rendered video URLs. Before installing or using it, consider: (1) only supply a NEMO_TOKEN if you trust the nemovideo.ai service; any token in your environment could be used to run export/upload requests. (2) Avoid uploading private or sensitive images you wouldn't want sent to a third-party GPU render service. (3) The skill will generate an anonymous token and make network calls if NEMO_TOKEN is absent — expect traffic to mega-api-prod.nemovideo.ai. (4) The metadata mentions a config path (~/.config/nemovideo/) though the runtime steps don't require it; that is minor but you may want to verify the skill won't access other local config files before granting broader file access. If you need stronger assurance, ask the publisher for source code or an official homepage and confirm the API host is legitimate.
Review Dimensions
- Purpose & Capability
- okThe name/description (image→video rendering) aligns with the required env var (NEMO_TOKEN) and the documented API endpoints on mega-api-prod.nemovideo.ai. The metadata's configPaths entry (~/.config/nemovideo/) is present but not used in the SKILL.md instructions — a minor inconsistency but not disproportional to the stated purpose.
- Instruction Scope
- noteSKILL.md gives explicit API calls for session creation, SSE-based generation, uploads, status, credits, and export. These network calls and local file uploads (multipart file posts) are expected for the stated service. Note: the instructions will cause the agent to read user-supplied local files for upload and to generate/retain a short-lived session token; this is normal for an upload/remote-render flow but worth user attention.
- Install Mechanism
- okNo install spec and no code files — instruction-only. That minimizes on-disk risk; nothing is fetched or executed by an installer.
- Credentials
- okOnly a single credential is required (NEMO_TOKEN) and the SKILL.md provides a clear anonymous-token fallback flow if the env var is absent. There are no unrelated credentials or broad system secrets requested.
- Persistence & Privilege
- okThe skill does not request 'always: true' or other elevated persistent privileges, nor does it instruct modification of other skills or global agent settings. Session tokens are used for the service and are ephemeral per the instructions.