The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides configuration instructions for AI agents to execute the @modelcontextprotocol/mcp-remote proxy via npx to connect to the remote Atlassian server.
[EXTERNAL_DOWNLOADS]: The installation process involves downloading the @modelcontextprotocol/mcp-remote package from the official npm registry. This package is maintained by the Model Context Protocol organization and is the standard tool for remote MCP connections.
[INDIRECT_PROMPT_INJECTION]: The skill enables the agent to read content from external Jira issues, Confluence pages, and Compass components. These sources could potentially contain malicious instructions intended to influence the agent's behavior.
Ingestion points: Data retrieved from Jira (JQL search), Confluence (CQL search), and Compass components in SKILL.md.
Boundary markers: The skill does not define specific delimiters for processed data, relying on the agent's standard tool-handling logic.
Capability inventory: The skill allows the agent to create and update content across Atlassian Cloud products using tools like jira_createIssue, confluence_updatePage, and compass_createComponent.
Sanitization: The skill relies on Atlassian's native permission model (OAuth 2.1/API tokens) and the AI agent's internal safety guardrails.
[CREDENTIALS_SAFE]: The documentation follows secure development practices by instructing users to use environment variables for sensitive data like ATLASSIAN_API_TOKEN instead of hardcoding credentials.

atlassian-mcp-server