The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs shell commands for curl and acli by interpolating variables such as site URLs, issue IDs, and Figma links derived from user input and API responses. The absence of explicit sanitization or escaping instructions for these variables presents a potential surface for command injection.\n- [DATA_EXFILTRATION]: Authentication credentials are retrieved via acli auth token and passed as Bearer tokens in network requests to the official atlassian.net domain. This is a standard operation for the well-known service, although passing secrets as command-line arguments involves a risk of exposure in system process logs.\n- [PROMPT_INJECTION]: The skill ingests untrusted data from user arguments and git branch names to automate ticket properties. It lacks boundary markers or instructions to disregard embedded commands, which is a vector for indirect prompt injection. \n
Ingestion points: User-provided instructions in $ARGUMENTS (Step 3) and local branch names via git branch (Step 2).\n
Boundary markers: None present in the instructions.\n
Capability inventory: Executes subprocesses using acli, git, and curl to manage files and network operations.\n
Sanitization: No validation or escaping of external content is specified before processing.\n- [EXTERNAL_DOWNLOADS]: The instructions describe how to install the acli tool using Homebrew from Atlassian's official distribution channel.

create-jira-ticket