Skillv1.0.0

ClawScan security

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

Scanner verdict

SuspiciousApr 13, 2026, 5:25 PM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill's declared purpose (convert Hindi text to AI-generated video) matches the network calls and single credential it requests, but several instruction-level details (automatic token generation/storing, suppression of raw API output, and filesystem/install-path probing) are unexplained and warrant caution.
Guidance: This skill appears to do what it says (calls nemo video APIs) and only asks for one credential (NEMO_TOKEN), but there are a few things to consider before installing: - Network endpoints: all API calls go to mega-api-prod.nemovideo.ai. If you have policy concerns about that domain, do not install. - Token handling: the skill will auto-request an anonymous token if NEMO_TOKEN is not present and suggests storing session_id/token for later use. If you prefer control, set NEMO_TOKEN yourself rather than letting the skill create one automatically, and periodically revoke tokens you don't recognize. - Filesystem probing: the instructions ask the agent to detect install paths (e.g., ~/.clawhub, ~/.cursor) and read frontmatter for attribution headers. That requires reading parts of your home directory — if you don’t want a skill to probe your filesystem, do not install or restrict its runtime permissions. - Reduced transparency: the SKILL.md tells the agent not to show raw API responses or token values to users. While this can be normal for UX, it also hides internal outputs; be cautious and monitor network/activity if you install. If possible, ask the skill author to clarify where session tokens are stored, why install-path detection is necessary, and to document any local file reads. If you proceed, prefer providing your own NEMO_TOKEN and review/rotate it regularly.

Review Dimensions

Purpose & Capability: noteThe skill's name/description align with the APIs it documents (video render, upload, export) and the single required env var (NEMO_TOKEN) is appropriate for a video-rendering service. Minor mismatch: the registry metadata listed no required config paths, but the SKILL.md frontmatter includes configPaths (~/.config/nemovideo/) — it's plausible but inconsistent.
Instruction Scope: concernRuntime instructions include network calls to https://mega-api-prod.nemovideo.ai for anonymous-token creation, session creation, SSE streaming, uploads and exports (expected). Concerns: the skill instructs the agent to 'detect install path' (probing ~/.clawhub, ~/.cursor, etc.) and to read the SKILL.md YAML frontmatter at runtime for attribution headers — both require local filesystem access beyond the purely networked video workflow. It also instructs the agent to suppress showing raw API responses or token values to the user, which reduces transparency and could hide unexpected behavior.
Install Mechanism: okNo install spec and no code files (instruction-only). This is lower-risk because nothing is downloaded or written by an installer.
Credentials: noteOnly NEMO_TOKEN is required, which is proportionate for an externally hosted video service. However, the frontmatter claims a config path (~/.config/nemovideo/) and instructions imply storing session_id/token for later requests — it's unclear where and how tokens/session IDs are persisted. The skill's instruction to auto-generate anonymous tokens (server-side returns token) is reasonable for convenience but increases the chance a token is created and stored without explicit user action.
Persistence & Privilege: okalways:false (not force-included) and no installation steps that alter other skills or system-wide settings. The only persistence implied is storing a session_id/token for subsequent API calls; the SKILL.md does not explicitly request system-wide config changes or other skills' credentials.