qinyan-citation

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local shell script scripts/search.sh to handle API interactions. The script includes basic input validation and uses standard system utilities.- [DATA_EXFILTRATION]: The skill performs network operations via curl to api.qinyanai.com. While this is required for the skill's functionality, the domain is not among the whitelisted repositories or services.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its processing of external academic metadata.
  • Ingestion points: Literature metadata (e.g., paper titles and abstracts) retrieved from external academic databases like Google Scholar and PubMed via scripts/search.sh.
  • Boundary markers: The skill lacks explicit boundary markers or instructions to the agent to disregard instructions potentially embedded within the retrieved literature data.
  • Capability inventory: The agent has access to general system tools and shell execution, which could be exploited if malicious content is processed from the search results.
  • Sanitization: There is no evidence of sanitization or filtering of the content returned by the external API before it is presented to the agent for citation formatting.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 08:22 AM
Security Audit — agent-trust-hub — qinyan-citation