扫描市场上的要约收购(Tender Offer)套利机会,分析价差、odd-lot优先权和风险,生成投资分析报告。
Security Analysis
high confidenceThe skill's stated purpose (scanning tender-offer arbitrage) aligns with its instructions and requirements: it only requires web access to public filings and market data, asks for no credentials, and saves reports locally.
Name/description match the runtime instructions: the SKILL.md instructs the agent to search SEC EDGAR and financial sites, verify filings, fetch market prices, compute spreads, analyze odd‑lot rules, rank opportunities, and write a report. There are no unexpected env vars, binaries, or config paths.
Instructions are specific and focused: explicit search queries, sites to check, data fields to extract, calculations to perform, risk checks, and an exact report format. The only broad capability required is network access (explicitly noted); the skill does not instruct reading unrelated local files, exfiltrating secrets, or performing trades.
No install spec and no code files — instruction-only. This minimizes on-disk execution risk; the skill relies on the agent's existing network/search capabilities.
The skill requests no environment variables, credentials, or config paths. Its needs (public web access and ability to save a report) are proportionate to the described financial scanning task.
The skill is not marked always:true and does not request persistent elevated privileges or modifications to other skills. It saves output to a local results directory as expected for a reporting skill.
Guidance
This skill appears coherent, but check these practical points before enabling it: (1) it needs network/search access — confirm your agent's network policies and that it will only query public data sources; (2) it does not request market-data or broker API keys, but accurate prices may require you to supply a read-only market-data API (use scoped, minimal-permission keys if needed); (3) ensure the agent is not granted brokerage credentials or the ability to place trades — the skill analyzes opportunities only; (4) review and sandbox where the agent writes reports (results/YYYY-MM-DD/report.md) so it cannot overwrite sensitive files; and (5) remember outputs are analytical and not investment advice.
Latest Release
v2.0.0
Initial release to ClawHub: scan, verify, and report arbitrage opportunities via SEC EDGAR using Python pipeline
Popular Skills
Published by @d-wwei on ClawHub