x9-agent-instructions

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script scripts/check_globals.py to ensure that specific blocks of text are synchronized across different configuration files. This script uses only the standard Python library.
  • [DATA_EXPOSURE]: The skill reads and modifies ~/.claude/CLAUDE.md and ~/.codex/AGENTS.md. While these are sensitive configuration files, this access is inherent to the skill's documented purpose of managing global agent instructions.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests content from external markdown files.
  • Ingestion points: The agent is instructed to read CLAUDE.md and AGENTS.md in SKILL.md, and the check_globals.py script reads these files directly from the home directory.
  • Boundary markers: The skill uses explicit comment markers (<!-- BEGIN SHARED PERSONAL CORE -->) to delimit sensitive content sections.
  • Capability inventory: The skill allows file reading/writing and execution of a local Python script.
  • Sanitization: No explicit sanitization or escaping of the ingested file content is performed before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 11:13 AM
Security Audit — agent-trust-hub — x9-agent-instructions