The Agent Skills Directory

[PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by instructing the agent to ingest and analyze untrusted data from YouTrack issues (summaries, descriptions, and comments).
Ingestion points: The 'Discovery Protocol' and 'Application' sections in SKILL.md involve querying and reviewing existing issues via API or tools.
Boundary markers: There are no instructions for using delimiters or negative constraints to prevent the agent from executing commands embedded in issue text.
Capability inventory: The agent is provided with a full command reference (references/commands.md) capable of modifying issue states, assignments, and visibility.
Sanitization: The skill does not provide methods for sanitizing external data before it is interpreted by the agent.
[EXTERNAL_DOWNLOADS]: The file .dev/reference-inventory.json contains references to official JetBrains documentation. These are well-known technology service domains and are considered safe sources for reference material.
[NO_CODE]: The skill is composed entirely of markdown documentation and a JSON configuration file; it contains no executable scripts or binary files.

youtrack