skills/ingpoc/skills/learn/Gen Agent Trust Hub

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 $ARGUMENTS and recent session context from the SKILL.md file.
  • Boundary markers: None present; content is interpolated directly into markdown templates.
  • Capability inventory: Uses Bash, Write, and Read tools to manage files in C:/Users/gurusharan.gupta/Agents/Claude Code/instincts.
  • Sanitization: Lacks explicit validation or escaping for the content being written to the filesystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 15, 2026, 09:46 AM
Security Audit — agent-trust-hub — learn