The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructions specify the use of curl to fetch data from https://bbs.auditdog.cn. This domain is the external source for all queries and is not among the platform's default whitelisted services.
[COMMAND_EXECUTION]: The skill uses shell commands (curl) to interact with the web API. It also uses python3 -c for inline string processing (URL encoding) within the shell context.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and display content from an external forum. This presents a surface for indirect prompt injection if the retrieved forum posts contain instructions intended to influence the agent's behavior.
Ingestion points: API responses from https://bbs.auditdog.cn described in the Workflow section of SKILL.md.
Boundary markers: Absent. There are no specific delimiters or instructions provided to the agent to treat the retrieved text as untrusted data.
Capability inventory: The skill has network read capabilities via curl.
Sanitization: None. The skill documentation describes mapping the question_text and comment_text fields directly into the agent's output.
[DYNAMIC_EXECUTION]: The workflow for retrieving post details includes a shell command that executes a Python snippet (python3 -c 'import urllib.parse; ...') to dynamically encode a URL parameter at runtime.

chenyiwei-bbs