summarize-hacker-news
Fail
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a shell command (
curlpiped topython3) to fetch and parse JSON data from the Hacker News API. Evidence:curl -L "https://hn.algolia.com/api/v1/search?tags=front_page" | python3 -c "...". - [EXTERNAL_DOWNLOADS]: The skill retrieves news metadata from
hn.algolia.com, which is a well-known service for Hacker News content. - [REMOTE_CODE_EXECUTION]: Automated scanners identified the piping of network data into a language interpreter. While the execution logic is hardcoded within the skill and used solely for parsing, this pattern is flagged for monitoring.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted third-party web content which is then summarized and sent via email. 1. Ingestion points: Content retrieved from articles via
tool read_webpage. 2. Boundary markers: Absent; there are no instructions to the agent to treat the fetched content as potentially malicious or to ignore embedded instructions. 3. Capability inventory: The skill utilizestool send_email_to_user, which could be manipulated if an attacker-controlled article successfully influences the agent's output. 4. Sanitization: No sanitization or validation of external content is performed before processing.
Recommendations
- HIGH: Downloads and executes remote code from: https://hn.algolia.com/api/v1/search?tags=front_page - DO NOT USE without thorough review
Audit Metadata