index-knowledge
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes standard shell commands (find, wc, awk, sed) to perform structural analysis and complexity scoring of the local codebase. These operations are restricted to directory and file metadata gathering.
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of ingesting and acting upon untrusted codebase data.
- Ingestion points: In SKILL.md, the workflow includes reading existing AGENTS.md files, scanning for entry points, and extracting project conventions and anti-patterns from source code comments.
- Boundary markers: The instructions do not specify any delimiters or warnings to the agent to ignore potentially malicious instructions embedded within the codebase being analyzed.
- Capability inventory: The skill has the capability to execute bash commands, write files to the disk (Write), and spawn multiple parallel subagents (Task) with prompts derived from the analyzed content.
- Sanitization: There is no evidence of filtering or sanitizing content retrieved from the codebase before it is used to populate prompts for subagents or the generated documentation.
Audit Metadata