Skillv1.0.0

ClawScan security

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

Scanner verdict

SuspiciousApr 13, 2026, 12:03 PM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill's declared credential (NEMO_TOKEN) and API workflow mostly match a cloud video renderer, but there are inconsistencies (declared config path in the skill text vs registry, and instructions to probe install/config locations) that warrant caution before installing.
Guidance: This skill largely behaves like a cloud video renderer — it will upload any images/clips you give it to an external API (mega-api-prod.nemovideo.ai) and uses a NEMO_TOKEN API token. Before installing or using: (1) Confirm the skill's source/author (homepage is missing); (2) avoid uploading sensitive or private images/video unless you trust the service; (3) ask the author how and where session tokens/anonymous tokens are stored (the SKILL mentions a config path but registry metadata did not); (4) be aware the skill asks the agent to inspect your home directories to determine 'X-Skill-Platform' — if you don't want local paths probed, ask for that to be removed; (5) prefer using a short‑lived/anonymous token rather than any long‑lived secret, and never supply unrelated credentials. If the author can clarify the config-path usage and remove unnecessary filesystem probing, the incoherence would be resolved and risk reduced.

Review Dimensions

Purpose & Capability: concernRequesting NEMO_TOKEN and calling a remote rendering API fits the described 'cloud video generator' purpose. However the SKILL.md frontmatter declares a required config path (~/.config/nemovideo/) and asks the agent to detect install paths for attribution headers — the registry metadata above this SKILL shows no required config paths. That mismatch is unexplained and not necessary for the core task.
Instruction Scope: concernRuntime instructions are mainly API-driven (session creation, SSE, upload, export) which is coherent. But the skill explicitly instructs the agent to detect the install path (~/.clawhub, ~/.cursor/skills/) and to read its own YAML frontmatter and (per frontmatter) a user config path. Probing user home directories or config locations is outside the minimal needs of generating/ uploading the user-supplied media and is privacy‑sensitive.
Install Mechanism: okThis is an instruction-only skill with no install spec and no code files; nothing is written to disk by an installer. That lowers supply-chain risk.
Credentials: noteOnly NEMO_TOKEN is declared as required which is proportional to calling a protected API. However SKILL.md describes generating an anonymous token and implies saving session_id and possibly storing tokens; combined with the frontmatter-config path (which the registry did not list) this suggests the skill may persist credentials/config locally — the storage location and justification are not clearly explained.
Persistence & Privilege: okThe skill does not request 'always:true' and uses normal autonomous invocation settings. There is no install-time behavior described that modifies other skills or system-wide settings.