The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local shell tools (curl and jq) to perform Jira Cloud REST API operations. It involves taking data retrieved from the Jira instance (such as option values and context names) and incorporating it into shell commands. The instructions do not specify sanitization or escaping of these values, creating a risk of command injection if the source data contains malicious shell metacharacters.
[DATA_EXFILTRATION]: The skill requires sensitive credentials including JIRA_API_TOKEN and JIRA_EMAIL. These credentials are used in curl commands sent to a variable JIRA_BASE_URL. Without validation of the base URL domain, there is a risk that these credentials could be sent to an attacker-controlled endpoint if the URL is misconfigured or manipulated via injection.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes untrusted data from an external Jira site.
Ingestion points: Data enters via jira_get_field_options and various Jira REST API GET endpoints used to fetch existing contexts and options.
Boundary markers: None. The skill does not define delimiters or provide instructions to the agent to treat the retrieved data as untrusted content.
Capability inventory: The skill has significant capabilities, including local shell access (curl, jq) and the ability to perform authenticated write operations (POST, PUT) on the Jira instance.
Sanitization: None detected. There are no steps provided to validate or escape external data before it is processed or used in subsequent API calls.

sigma-atlassian-toolkit