Back to skill
Skillv1.0.1
ClawScan security
Pod Ai Agent · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
SuspiciousApr 30, 2026, 9:55 AM
- Verdict
- suspicious
- Confidence
- medium
- Model
- gpt-5-mini
- Summary
- The skill is mostly coherent with its POD automation purpose and only needs a single vendor API key, but there is a metadata mismatch about required environment variables and some minor scope/metadata inconsistencies the user should verify before installing.
- Guidance
- This skill appears to be a documentation/guide that calls a Yunlv (TradeGPT) API and otherwise only provides playbooks and checklists. Before installing/providing secrets: 1) Confirm whether the platform actually requires TRADEGPT_API_KEY (the package files declare it but the registry summary omitted it). 2) If asked, only supply an API key you control for yunlvai.com and avoid giving shop/store platform credentials unless you explicitly intend to connect those services. 3) Understand that the skill will send strategy/tooling queries to https://api.yunlvai.com—don’t send sensitive customer data unless you accept that the vendor will receive it. 4) Keep API keys out of code, use platform secret storage, and rotate keys periodically. 5) Because this is instruction-only (no code), the main remaining risk is data sent to the vendor API and any downstream third-party integrations you configure; review yunlvai.com's privacy/security policy and test in a controlled environment first. Finally, ask the publisher to correct the metadata inconsistency about required env vars so you know exactly what will be requested.
Review Dimensions
- Purpose & Capability
- noteThe skill's name/description (POD AI agent) align with the runtime instructions and the included reference docs. The SKILL.md and clawhub.yaml declare a primaryEnv TRADEGPT_API_KEY and an API endpoint (https://api.yunlvai.com) which is sensible for a vendor-backed strategy/recommendation service. However, the registry summary at the top of the package incorrectly listed "Required env vars: none" while the bundle files declare TRADEGPT_API_KEY — an inconsistency worth confirming with the publisher.
- Instruction Scope
- okThis is an instruction-only skill (no code/install). SKILL.md is guidance for building/operating POD workflows and references local resource files and a local storage path (./data/yunlv-ski). It does not instruct the agent to read arbitrary system files or to exfiltrate unspecified secrets; it recommends integrating third‑party tools and storing API keys appropriately. No instructions claim to access user store platforms (the metadata explicitly says the Yunlv API does not access shop platforms).
- Install Mechanism
- okNo install spec and no code files are present; the skill is instruction-only. That minimizes disk-write/execution risk.
- Credentials
- noteThe package declares a single primary credential TRADEGPT_API_KEY (used to call yunlvai's TradeGPT API). A single vendor API key is proportionate to the skill's stated purpose. The inconsistency is that the registry header showed "none" for required env vars while clawhub.yaml and SKILL.md require TRADEGPT_API_KEY — verify which metadata the platform will enforce and whether you must provide that key.
- Persistence & Privilege
- okalways:false and disable-model-invocation:false (normal). The skill mentions storing data under a relative path (./data/yunlv-ski) which suggests local caching but does not request system-wide privileges or modify other skills/config. No elevated or persistent privileges are requested.