The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-submitted content (topics and drafts) in Phase 1 and Phase 3 of its workflow. This represents a data ingestion surface where external input is analyzed by the agent.
Ingestion points: The Phase 1: 接收内容 section in SKILL.md prompts for user topics and draft content.
Boundary markers: The skill uses structured reporting templates and headers to separate its logic from user data, though it does not include explicit 'ignore instructions' warnings.
Capability inventory: The skill lacks dangerous capabilities; it does not perform network operations, file writing, or command execution.
Sanitization: No explicit sanitization of user input is implemented within the instructions.
[NO_CODE]: The skill consists entirely of markdown instructions and YAML configuration. There are no scripts, binaries, or executable files included in the package.

dbs-content