dir-tree-index
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted file content to generate summaries for STRUCTURE.md files.
- Ingestion points: Reads the first 30 lines of files within the target directory to generate summaries in SKILL.md Step 3.
- Boundary markers: None. The skill does not use delimiters or instructions to ignore embedded commands within the files being summarized.
- Capability inventory: Uses find via shell, reads file contents, and writes STRUCTURE.md files back to the repository.
- Sanitization: There is no evidence of sanitization or filtering of the content read from files before it is interpolated into the summary.
- [COMMAND_EXECUTION]: The skill uses bash to execute find commands to discover directories and files in Step 1 and Step 2.
- [DYNAMIC_EXECUTION]: The skill constructs shell commands by substituting a user-provided target directory into a template (find ...). If the target directory name is not properly sanitized by the agent platform, this could lead to command injection.
Audit Metadata