Skills
skills/hitoshura25/claude-devtools/version-management/Gen Agent Trust Hub

version-management

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses shell commands to create and make scripts executable.\n
  • Evidence: The skill executes chmod +x on generated scripts scripts/version-manager.sh and scripts/gradle-version.sh.\n
  • Context: These are standard operations for installing local utility scripts and are performed on files authored by the skill itself.\n- [PROMPT_INJECTION]: The version management logic is exposed to indirect prompt injection via repository tags.\n
  • Ingestion points: The get_latest_version function in version-manager.sh reads external data using git tag -l.\n
  • Boundary markers: No specific delimiters or warnings are used to wrap the version data.\n
  • Capability inventory: The skill can write to the filesystem (version.properties) and set GitHub Action outputs via GITHUB_OUTPUT.\n
  • Sanitization: Version strings are parsed using IFS='.' and sed, providing structural validation but no explicit shell escaping.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 12:40 PM