The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external sources such as "stakeholder's documents", code repositories, and local files. This creates an attack surface for indirect prompt injection if those sources contain malicious instructions.
Ingestion points: Processes content from stakeholder documents, existing artifacts (dashboards, scripts, mocks), and repository files accessed via origin/<default>.
Boundary markers: The instructions do not define specific delimiters or "ignore embedded instructions" warnings for the external content it processes.
Capability inventory: The skill can perform environment lookups, read files, analyze code paths, and dispatch sub-agents for tasks.
Sanitization: No explicit sanitization, escaping, or filtering of external content is defined in the instructions.
[COMMAND_EXECUTION]: The instructions direct the agent to perform environment lookups to find facts, specifically mentioning repository status checks using origin/<default> after a fetch. This involves executing commands to interact with the version control system.
[EXTERNAL_DOWNLOADS]: The skill utilizes git fetch to synchronize with remote repositories (origin) to ensure the agent analyzes the authoritative state of the project rather than a potentially stale local cache. This is a standard and expected operation for a developer-oriented analysis tool.

grilling