The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external content such as WeChat articles or social media posts provided by the user or retrieved via auxiliary tools. This introduces a potential surface for indirect prompt injection if the processed content contains hidden instructions meant to manipulate the agent's logic.
Ingestion points: External content is ingested when analyzing user-provided articles or fetching data via baokuan-article-analysis and global-content-search (referenced in SKILL.md).
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' warnings for the external data being processed.
Capability inventory: The skill itself is instructional markdown and lacks direct executable code, file-writing, or network capabilities, which limits the impact of an injection.
Sanitization: There are no explicit instructions for the agent to sanitize or filter the content of articles before analysis.

gzh-positioning