The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes several standard development tools such as gh, npm, bundle, and rails to perform environment analysis and version checking. These operations are aligned with the stated purpose of assessing a Rails application for upgrades.
[EXTERNAL_DOWNLOADS]: The skill retrieves information from well-known and reputable sources, including official Rails documentation (guides.rubyonrails.org), community-standard diff tools (railsdiff.org), and the GitHub API for the official Rails repository. These sources are considered safe within the context of Rails development.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it ingests data from local project files.
Ingestion points: Reads local files including Gemfile, Gemfile.lock, package.json, config/importmap.rb, and Rails configuration files (SKILL.md Steps 1, 2, 7.2, 7.4, 9.1).
Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore potentially malicious content within these files.
Capability inventory: The skill has capabilities including shell command execution (gh, npm, bundle, rails) and network requests via WebFetch and gh CLI (SKILL.md Steps 3, 5, 6, 7.3, 9.4).
Sanitization: No validation or sanitization is performed on the content read from the local files before processing.

skill-rails-upgrade