explain-artifacts
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and summarizes untrusted external artifacts which could contain malicious instructions or script payloads.\n
- Ingestion points: The agent is instructed to identify and inspect relevant artifacts from the request and conversation (SKILL.md).\n
- Boundary markers: The skill includes explicit instructions to 'Escape artifact text inserted into HTML' and 'Treat artifact markup as text rather than copying it as executable HTML' (SKILL.md).\n
- Capability inventory: The agent can create temporary directories, write files, and execute shell commands to open them (SKILL.md).\n
- Sanitization: Instructions mandate that the agent must escape all text derived from artifacts before placing it in the HTML template.\n- [COMMAND_EXECUTION]: The skill uses system commands to manage the environment and display the final output to the user.\n
- Evidence: Mentions of
mktemp -dfor temporary directory creation and the use ofopenorxdg-opento launch the resulting HTML file in a browser (SKILL.md).
Audit Metadata