Back to skill
Skillv1.0.0
ClawScan security
预测市场套利 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignApr 2, 2026, 1:15 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill's code, runtime instructions, and requested environment variable (AISA_API_KEY) are consistent with its stated purpose of read-only arbitrage scanning via the AIsa API.
- Guidance
- This skill appears to do what it says: read-only market queries to AIsa and local arbitrage calculations. Before installing: 1) Treat your AISA_API_KEY like a billing credential — limit its scope if possible and monitor usage to avoid unexpected charges. 2) Do not supply private keys or wallet secrets — the scripts only need public IDs/addresses. 3) Be aware the client can query wallet activity/positions for given addresses (public data), so avoid providing sensitive identifiers you don't want queried. 4) Review the bundled Python scripts yourself or run them in a sandbox to verify behavior and expected network calls. If you rely on this in production, confirm you trust the AIsa provider and understand the per-request cost.
Review Dimensions
- Purpose & Capability
- okName/description match the implementation: Python scripts call https://api.aisa.one to find matching markets, compare prices, and check orderbooks for Polymarket and Kalshi. Requested binary (python3) and AISA_API_KEY credential are appropriate for this purpose.
- Instruction Scope
- noteSKILL.md instructs only read-only queries and local calculation, and the code uses only HTTPS GETs to api.aisa.one. The code also exposes read-only endpoints for wallet/activity/positions (i.e., can query public wallet addresses) — this is in-scope for market analysis but is worth noting because these calls can reveal public activity for specified wallet addresses if the user requests them.
- Install Mechanism
- okNo install spec; skill is instruction + bundled Python scripts only. Nothing is downloaded or written during install — lowest-risk delivery model.
- Credentials
- okOnly AISA_API_KEY is required and declared as the primary credential, which is proportional for a service-backed data client. Note: AIsa is a paid API (SKILL.md says ~$0.01/query), so the key can incur costs if abused or used excessively.
- Persistence & Privilege
- okalways:false (no forced inclusion). disable-model-invocation is false (normal) — the skill can be invoked by the agent autonomously, which is expected behavior. The skill does not request system-wide config or modify other skills.
