learn
Pass
Audited by Gen Agent Trust Hub on Jun 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a simple shell command (
ls) to list files in the local instincts directory. This is used for checking duplicates and is within the expected scope of the tool. - [INDIRECT_PROMPT_INJECTION]: The skill creates files based on session context and user input, which presents a surface for indirect prompt injection.
- Ingestion points: Processes
$ARGUMENTSand recent session context from theSKILL.mdfile. - Boundary markers: None present; content is interpolated directly into markdown templates.
- Capability inventory: Uses
Bash,Write, andReadtools to manage files inC:/Users/gurusharan.gupta/Agents/Claude Code/instincts. - Sanitization: Lacks explicit validation or escaping for the content being written to the filesystem.
Audit Metadata