The Agent Skills Directory

[DATA_EXFILTRATION]: The script scripts/fetch_growth_rank.py contains logic to automatically read sensitive shell configuration files (~/.bashrc, ~/.zshrc, ~/.bash_profile, ~/.profile) to extract the REDFOX_API_KEY. Accessing these files is a security concern as they typically contain other sensitive environment variables and system configurations. The extracted key is then transmitted to the vendor's API at redfox.hk.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it fetches data (article titles, summaries, and content) from an external API and instructs the agent to analyze it without sanitization or clear boundary markers. This could allow malicious content from the data source to influence the agent's behavior.
Ingestion points: Data is ingested via scripts/fetch_growth_rank.py from the redfox.hk API.
Boundary markers: No delimiters or 'ignore embedded instructions' warnings are present to separate the fetched data from the agent's internal instructions.
Capability inventory: The skill uses scripts/fetch_growth_rank.py which performs network operations and the agent is instructed to perform complex content analysis and summarization.
Sanitization: There is no evidence of sanitization, escaping, or validation of the external content before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the requests library via pip, which is a standard and well-known dependency for Python-based network operations.

wechat-fastest-growing