The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires downloading its core logic from an external, untrusted source at https://github.com/doane2002cn/moontv-openclaw-skill.git. This repository is not recognized as a trusted organization or a verified resource belonging to the skill author.
[COMMAND_EXECUTION]: The skill instructs the user/agent to execute multiple shell and Python commands, including git clone, python moontv_daily.py, and pytest. These commands run code fetched from an external repository, which could be modified by a third party to include malicious payloads.
[INDIRECT_PROMPT_INJECTION]: The skill has a vulnerability surface for indirect prompt injection attacks.
Ingestion points: Data is ingested from hundreds of third-party CMS sources via API calls in moontv_daily.py using fetch_cms_data.
Boundary markers: None. The prompt template in highlight_generator.py uses a standard JSON structure to present data to the LLM but does not include explicit instructions to ignore potentially malicious commands embedded in the synopsis text.
Capability inventory: The skill can perform network operations (requests) and file system operations (writing reports to output/ and deleting files via cleanup_old_reports).
Sanitization: The skill only truncates the synopsis to 200 characters; it does not perform any validation, filtering, or escaping of the content before it is processed by the LLM.

moontv-openclaw-skill