skills/johnlindquist/claude/packx/Gen Agent Trust Hub

packx

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill uses authoritative markers to override default agent behavior regarding CLI flags. Evidence: Explicit use of "CRITICAL", "MANDATORY", and "non-negotiable" to force the inclusion of the --limit 49k flag in commands.
  • [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest local file content, which may contain instructions that could influence downstream AI models. 1. Ingestion points: Reads local source files through direct paths and glob patterns defined in SKILL.md. 2. Boundary markers: Supported output formats include XML, Markdown, and JSONL to delineate file content. 3. Capability inventory: Uses clipboard access via the -c flag and file writing via standard shell redirection or output flags. 4. Sanitization: No automated sanitization or filtering of file content is specified.
  • [EXTERNAL_DOWNLOADS]: Instructions provide for the installation of an external global NPM package. Evidence: npm install -g packx command provided in the Prerequisites section.
  • [COMMAND_EXECUTION]: Provides numerous examples of shell command execution for filesystem and git operations. Evidence: CLI usage examples for file searching, glob filtering, and git diff calculations.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 08:25 PM
Security Audit — agent-trust-hub — packx