The Agent Skills Directory

[CREDENTIALS_UNSAFE]: Both open-link.py and web-search.py contain a hardcoded application key (38d2391985e2369a5fb8227d8e6cd5e5) used to generate MD5 signatures for API authentication headers.- [COMMAND_EXECUTION]: The skill requires the agent to execute shell commands to run the Python scripts (python web-search.py and python open-link.py) using arguments derived from user queries and external search results.- [EXTERNAL_DOWNLOADS]: The skill performs network requests to an external API endpoint (https://autoglm-api.zhipuai.cn) and a local service (http://127.0.0.1:53699/get_token) to retrieve data and authentication tokens.- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted data from the web.
Ingestion points: Content is retrieved from arbitrary external URLs via open-link.py and search snippets via web-search.py (as described in SKILL.md).
Boundary markers: The skill instructions lack delimiters or explicit directives for the agent to ignore instructions embedded within the retrieved web content.
Capability inventory: The skill provides capabilities to execute shell commands and perform further network requests via the included Python scripts.
Sanitization: There is no evidence of content filtering or sanitization of the retrieved web text before it is processed by the agent.

autoglm-deepresearch