The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The documentation provides standard instructions to install the mcp-atlassian package for Jira and Confluence integration. This package is from a well-known service provider.
[COMMAND_EXECUTION]: The skill suggests running the Atlassian MCP server using uvx, a standard and secure tool for executing Python-based CLI applications.
[PROMPT_INJECTION]: The skill ingests untrusted data in the form of user stories and acceptance criteria (SKILL.md) and passes them to sub-agents via the TaskCreate capability. While this creates an indirect prompt injection surface with no explicit boundary markers or sanitization, it is the primary intended functionality of the skill.
[SAFE]: Detailed instructions for configuring Jira API tokens recommend using local configuration files, which is consistent with standard secret management practices. No malicious code, exfiltration, or persistence mechanisms were found.

agile