repo-research-analyst

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as its primary function involves reading and summarizing untrusted data from external repositories, including source code, issue trackers, and documentation.
  • Ingestion points: The skill uses file-reading and content-searching tools (Glob, Grep, Read) to process arbitrary files within a repository (SKILL.md).
  • Boundary markers: The instructions do not define clear delimiters or "ignore previous instructions" safety markers for the content being analyzed.
  • Capability inventory: The agent has access to the file system via multiple tools and the capability to execute shell commands (SKILL.md).
  • Sanitization: No sanitization or validation of the ingested repository content is performed before processing, allowing potential instructions in the source code to influence agent behavior.
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the shell to execute ast-grep and other commands when no native tool equivalent exists. While scoped to the research task, this provides a mechanism for command execution within the environment (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 08:22 PM