Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent RSS briefing skill; its web fetching and local history log are expected for the stated purpose, though users should review those behaviors before running it.
Install this only if you are comfortable with the agent contacting configured RSS feeds and article links and maintaining a local pushed_history.log file. Review the RSS configuration before use, and avoid enabling command-line or write access unless you want deduplication history to be updated.
The trigger keywords are broad and generic, including common terms like 'rss', '链接', and 'aggregator', which raises the chance of unintended activation. In this skill, accidental invocation is more concerning because execution includes external content fetching and local file modification via pushed_history.log.
The repeated keyword section again uses vague activation phrases without clear constraints, reinforcing the risk of false-positive invocation. Because the skill reads configured RSS sources, fetches linked content, and appends to a local log, an unintended trigger can cause unwanted external access and state changes.
The skill instructs fetching RSS feeds and reading linked article content, but does not clearly warn users that it will access external network resources. This matters because external fetches can expose metadata, interact with untrusted content, and trigger unintended requests in restricted or sensitive environments.
The workflow instructs the agent to append items to a local pushed_history.log file but does not warn the user that local state will be modified. Silent persistence can surprise users, create auditability problems, and in shared environments may alter files without meaningful consent.
The API Key section states '无需额外API Key', but the dependency table immediately above lists 'LLM API' as required. While this may mean the host agent supplies credentials, the wording is misleading and can cause operators to misunderstand what external capability the skill relies on.
No suspicious patterns detected.
