The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it retrieves and processes content from external Atlassian sources.
Ingestion points: Fetches issue summaries, descriptions, and comments from Jira, and page content from Confluence.
Boundary markers: No isolation delimiters or instructions are used to prevent the agent from obeying instructions embedded in retrieved data.
Capability inventory: The skill includes the ability to create, update, and delete Jira issues and Confluence pages.
Sanitization: Data retrieved from the Atlassian API is not sanitized before being presented to the agent.
[COMMAND_EXECUTION]: Shell commands such as curl and jq are used correctly to facilitate API requests and response parsing.
[CREDENTIALS_UNSAFE]: Sensitive information is properly managed through environment variables and the vm0_secrets manifest, avoiding hardcoded credentials.
[DATA_EXFILTRATION]: Network traffic is directed exclusively to official Atlassian Cloud domains (atlassian.net), which are recognized as well-known and trusted services for this integration.

atlassian