coding-global-memory

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill triggers the execution of mkdir -p to create storage directories. The directory name is derived by sanitizing the current working directory path (replacing slashes with dashes), which reduces risk but still involves shell-level directory creation.
  • [PROMPT_INJECTION]: The skill is vulnerable to Indirect Prompt Injection (Category 8) because it ingests untrusted data and persists it for later retrieval.
  • Ingestion points: The skill scans local project files (e.g., package.json, go.mod) and summarizes the current conversation history to generate memory content.
  • Boundary markers: No boundary markers or 'ignore' instructions are specified for the data being summarized or read from existing memory files.
  • Capability inventory: The skill possesses file system read/write capabilities and the ability to execute directory creation commands.
  • Sanitization: There is no mention of sanitizing or filtering instructions that might be embedded in the project files or conversation before they are written to the persistent MEMORY.md files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 06:49 AM