skills/dracohu2025-cloud/draco-skills-collection/daily-ai-agent-aigc-top-news-dingtalk/Gen Agent Trust Hub
daily-ai-agent-aigc-top-news-dingtalk
Pass
Audited by Gen Agent Trust Hub on May 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses several system commands and external CLI tools to perform its tasks.
- Evidence in
SKILL.md: Executesdatefor timezone verification,gitfor local repository checks, anddws(DingTalk Command Line Tool) for document creation and retrieval. - [EXTERNAL_DOWNLOADS]: The skill retrieves data from external APIs to aggregate news content.
- Evidence in
SKILL.md: Uses Python'surllib.requestto fetch release and pull request data from the GitHub API (api.github.com). - [REMOTE_CODE_EXECUTION]: The skill executes Python scripts to process data and interact with external services.
- Evidence in
SKILL.md: Runs inline Python code blocks and executes external scripts such asnews_aggregator_run.pyandsync_doc_to_dingtable.pyusingpython3. - [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it processes untrusted content from the web.
- Ingestion points: News candidates are fetched from Hacker News, GitHub, Hugging Face, and various official sites (Phase 2 and 3 in
SKILL.md). - Boundary markers: The skill does not explicitly mention delimiters or instructions to ignore embedded prompts in the ingested data.
- Capability inventory: The skill has the ability to write to DingTalk documents and execute local shell commands.
- Sanitization: No specific sanitization or filtering of the ingested news content is mentioned before it is processed by the agent to generate the briefing.
Audit Metadata