The Agent Skills Directory

[COMMAND_EXECUTION]: The scanner utility in scripts/utils/changed_files.rb executes git diff using shell backticks to identify modified files in the repository. This is an expected and necessary behavior for a developer tool designed to perform incremental analysis on a codebase.
[INDIRECT_PROMPT_INJECTION]: The skill acts as a static analysis engine that ingests untrusted data in the form of project source code and templates through Scanner#read_file.
Ingestion points: Project-specific files (.rb, .erb, .haml, .slim) are read into memory to be scanned for PII patterns.
Boundary markers: The scanner does not currently use explicit delimiters or "ignore instructions" markers when outputting findings, though it operates as a tool providing structured JSON results.
Capability inventory: The scanner has capabilities to read local project files and execute local git commands.
Sanitization: Content is processed via regular expressions for pattern matching; while snippets are returned in the analysis, they are not executed or evaluated as code within the skill's own logic.
[SAFE]: All external URLs and dependencies referenced in the documentation, such as official Ruby on Rails guides, GDPR/LGPD legal texts, and the pdscan utility by known community contributor ankane, are from trusted or well-known organizations and represent standard industry resources.

privacy-by-design-rails