Skillv1.0.0

ClawScan security

Pika Ai Video Generation · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 26, 2026, 4:42 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill mostly matches a video-generation integration but has small inconsistencies (metadata vs registry), instructs the agent to auto-create and store short‑lived API tokens, and will upload user files to an external service — review before installing or uploading sensitive content.
Guidance: This skill appears to be a reasonable wrapper for a cloud video-rendering API, but review the following before installing or using it: 1) It will upload whatever files you give it (videos, images, audio) to the external host (mega-api-prod.nemovideo.ai). Do not upload private or sensitive files unless you trust the service and its privacy policy. 2) If you don't provide a NEMO_TOKEN, the skill will automatically request an anonymous token for you and treat that as the credential — understand what that token allows (100 free credits, 7‑day expiry) and where it will be stored. 3) There is an inconsistency: the skill metadata references a config path (~/.config/nemovideo/) even though registry metadata lists no config paths — ask the publisher which is correct and whether the skill will read or write files on disk. 4) Confirm the API domain and the service's reputation (search for 'nemovideo' / 'nemo video' official docs) if you plan to send sensitive content or rely on billing. If you want tighter control, prefer supplying your own NEMO_TOKEN obtained directly from the provider and avoid letting the agent auto-provision credentials.

Review Dimensions

Purpose & Capability: noteName, description, required env var (NEMO_TOKEN) and the REST endpoints in SKILL.md align with a cloud video-rendering service. However the YAML frontmatter includes a configPaths entry (~/.config/nemovideo/) while the registry metadata lists no required config paths — this mismatch is an incoherence in declared requirements. The skill also derives attribution headers from install paths (reads install location) which implies some filesystem-awareness beyond simply calling an API.
Instruction Scope: noteSKILL.md keeps to the stated purpose: it instructs the agent to create sessions, upload media, stream SSE events, request renders, and poll results at the stated API host. Two items to note: (1) if NEMO_TOKEN is absent the instructions automatically POST to an anonymous-token endpoint and treat the returned token as NEMO_TOKEN (the skill will obtain credentials on the user's behalf); (2) the skill will upload user-provided files (up to ~200MB) to the external domain (mega-api-prod.nemovideo.ai) — expected for this functionality but relevant for privacy/data-exfiltration risk. The document also instructs the agent not to display raw API responses or token values, which is a UX guideline but reduces transparency about what the agent stores or transmits.
Install Mechanism: okThis is an instruction-only skill with no install spec and no packaged code to download or execute, which is the lowest-risk install profile.
Credentials: okOnly a single credential (NEMO_TOKEN) is declared as required and is appropriate for a third‑party API integration. The skill's instructions for creating an anonymous token when none is present are proportionate to the described free-trial flow (100 free credits, 7 days).
Persistence & Privilege: noteThe skill is not always-enabled and does not request elevated platform privileges. It does instruct storing and reusing a session_id and/or token for subsequent requests; it may also read an install path to set attribution headers. The earlier-mentioned metadata entry for ~/.config/nemovideo/ suggests possible filesystem read/write intent — inconsistent with registry metadata and worth clarification.