Skillv1.0.0

ClawScan security

Image To Video Kaise Banaye · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 23, 2026, 4:38 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill's stated purpose (convert images to short videos) matches the runtime instructions and required NEMO_TOKEN, but there are a few inconsistencies and privacy/ownership risks (unknown backend, metadata mismatch) you should understand before installing.
Guidance: This skill appears to do what it says — it uploads your images to an external service (mega-api-prod.nemovideo.ai) and returns rendered MP4s. Before installing or using it: (1) Verify the provider (there's no homepage or publisher info); (2) Do not upload sensitive images — uploads go to an external, unverified host; (3) Only provide a NEMO_TOKEN that is scoped/dedicated to this service (do not reuse long-lived tokens you use elsewhere); (4) Ask the author to clarify the config path discrepancy (~/.config/nemovideo/ vs. 'none' in registry metadata) and for a privacy/retention policy for uploaded media; (5) If you want to test, try non-sensitive images first and confirm where downloads are hosted and how long files are retained. These steps will reduce privacy and trust risks.

Review Dimensions

Purpose & Capability: noteName/description match the instructions: the skill uploads images and calls an external video-rendering API. Requesting a single NEMO_TOKEN credential is consistent with a third‑party media API. However, the SKILL.md frontmatter lists a config path (~/.config/nemovideo/) while the registry metadata reported 'required config paths: none' — that's an internal inconsistency worth clarifying.
Instruction Scope: concernThe instructions tell the agent to upload user-supplied images and poll/stream results from https://mega-api-prod.nemovideo.ai and to include Authorization: Bearer <NEMO_TOKEN>. That is expected for a cloud render service but means user images and any supplied token will be transmitted to an external, unknown backend. The skill also instructs acquiring an anonymous token if no NEMO_TOKEN is present, which requires generating a UUID and POSTing to the same unknown endpoint. No instructions ask the agent to read unrelated files or other env vars, but the explicit upload of potentially sensitive images to an unverified third party is a privacy risk.
Install Mechanism: okInstruction-only skill with no install spec and no code files — lowest disk/write risk. There is no package download or archive extraction.
Credentials: noteOnly one environment variable is required (NEMO_TOKEN), which is proportionate for a service that authenticates API calls. Caveats: the SKILL.md frontmatter implies a config path (~/.config/nemovideo/) and the registry metadata did not list it — inconsistent. The skill will use any NEMO_TOKEN found in the environment; ensure that token is dedicated to this service and not reused for other services to avoid accidental exfiltration.
Persistence & Privilege: okalways:false (normal). The skill is user-invocable and can be invoked autonomously (default), which is expected. It does not request to modify other skills or system-wide settings. Note: autonomous invocation combined with networked upload capability increases blast radius if the skill were malicious — consider this in your risk decision.