Back to skill
Skillv1.0.0
ClawScan security
Editor Ai For Beginners · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
SuspiciousApr 10, 2026, 11:43 PM
- Verdict
- suspicious
- Confidence
- medium
- Model
- gpt-5-mini
- Summary
- The skill's declared purpose (cloud video editing via a Nemo API) aligns with most requirements, but the runtime instructions ask the agent to read local file paths and detect install/config locations and there is a small metadata mismatch — this creates scope‑creep risk before you trust it with files or a token.
- Guidance
- This skill appears to be a thin client for a Nemo cloud video editing API and mostly coherent, but exercise caution before installing or using it with sensitive data. Specifically: (1) don't provide your NEMO_TOKEN unless you trust the backend — the skill will use it as a bearer token for all requests; (2) avoid uploading sensitive local files — the instructions include reading files from arbitrary paths (files=@/path), so prefer attaching files via the chat UI or testing with non-sensitive samples first; (3) verify the API domain (mega-api-prod.nemovideo.ai) and the service's privacy/data-retention policies if you plan to upload real content; (4) note the SKILL.md references local config paths and install-path detection — if you want to limit exposure, run the skill in a restricted environment or deny access to your filesystem; (5) if you don't have a NEMO_TOKEN, the skill will request an anonymous token from the remote API automatically — be aware that this still transmits metadata (X-Skill-* headers) to the provider. If these behaviors are acceptable, test with throwaway tokens and non-sensitive clips first. If you need more certainty, ask the skill author for a privacy/security policy and clarification about local filesystem access and what the attribution headers contain.
Review Dimensions
- Purpose & Capability
- noteName/description match the actions in SKILL.md: it calls a remote Nemo video API and needs a NEMO_TOKEN. That is coherent. However the SKILL.md metadata references a config path (~/.config/nemovideo/) while the registry metadata lists no required config paths — minor inconsistency in declared vs. actual metadata.
- Instruction Scope
- concernThe instructions direct the agent to create sessions, call SSE endpoints, and upload videos. Crucially, upload examples include multipart 'files=@/path' syntax and the doc tells the agent to detect install paths (e.g. ~/.clawhub/) to set X-Skill-Platform. That implies the agent may be expected to read arbitrary local filesystem paths rather than only using user-attached files; this is scope creep and a privacy/risk concern. The skill also instructs obtaining an anonymous token if none is present (POST to external API) which is expected but means network calls will be made automatically.
- Install Mechanism
- okThere is no install spec and no code files — instruction-only skill. That reduces risk compared with arbitrary downloaded code.
- Credentials
- noteOnly one credential (NEMO_TOKEN) is declared as required, which is appropriate for a cloud-editing integration. The SKILL.md also documents a flow to obtain a short-lived anonymous token if NEMO_TOKEN is absent — reasonable but be aware that the skill will call the external auth endpoint and use that token for subsequent requests. The presence of metadata-reported configPaths (in SKILL.md) is not reflected in the registry and could imply the skill expects access to a local config directory.
- Persistence & Privilege
- okalways:false and no install or persistent permissions are requested. The skill does suggest detecting install/config paths, but it does not request to be always-enabled or to modify other skills; no elevated persistence is requested.