memory-optimize

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill accesses local Claude configuration and memory files (e.g., ~/.claude/projects/**/memory/*.md, ~/.claude/CLAUDE.md). This access is localized to the agent's own environment for the purpose of optimization.\n- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to perform file system operations and environment checks, which is explicitly allowed in the skill configuration.\n- [PROMPT_INJECTION]: The skill processes potentially untrusted content from history and memory files, creating a surface for indirect prompt injection.\n
  • Ingestion points: Reads memory content from markdown files in ~/.claude/projects/ (SKILL.md).\n
  • Boundary markers: Absent. No specific boundary markers or instructions to ignore embedded instructions are provided.\n
  • Capability inventory: Utilizes Read, Write, Edit, Glob, Grep, Bash, Task, and AskUserQuestion tools (SKILL.md).\n
  • Sanitization: Absent. No explicit sanitization or validation of memory content is performed.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 08:20 AM