The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content (Jira tickets) provided via user arguments or pasted context. A maliciously crafted Jira ticket could attempt to manipulate the agent into performing unauthorized actions.
Ingestion points: User-provided Jira tickets via $ARGUMENTS or direct paste in the conversation context.
Boundary markers: Absent; the instructions do not include delimiters or warnings to ignore commands embedded within the external ticket data.
Capability inventory: The skill leverages tools for file system access, code modification, shell command execution, and Git version control.
Sanitization: Absent; there is no logic described for validating or filtering the content of the ticket before processing.
[COMMAND_EXECUTION]: The skill executes local shell commands using pnpm for project verification steps (linting, testing, and type checking).
Evidence: Step 5 executes pnpm typecheck, pnpm lint, and pnpm test:run. These are standard development workflows and do not involve remote script execution or privilege escalation.

jira