The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the addition of an external Model Context Protocol (MCP) server endpoint: https://rube.app/mcp. This domain is associated with Composio, a known provider of agentic tools, and aligns with the skill author's professional context. The setup uses standard MCP configuration patterns and relies on Jira OAuth via RUBE_MANAGE_CONNECTIONS for secure credential handling rather than hardcoded keys.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it ingests untrusted data from Jira.
Ingestion points: Tools such as JIRA_GET_ISSUE, JIRA_SEARCH_FOR_ISSUES_USING_JQL_POST, and JIRA_LIST_ISSUE_COMMENTS retrieve content (issue descriptions, comments, and summaries) that can be modified by external users.
Boundary markers: The instructions do not specify any delimiter or boundary markers to separate retrieved Jira content from the agent's instructions.
Capability inventory: The skill includes high-privilege write capabilities, including JIRA_EDIT_ISSUE, JIRA_ASSIGN_ISSUE, and JIRA_ADD_USERS_TO_PROJECT_ROLE, which could be abused if an agent follows instructions embedded within a Jira issue.
Sanitization: There are no explicit sanitization or filtering steps mentioned for processing external content before interpolation into the prompt.
Note: This risk is inherent to the skill's primary purpose of Jira management and is mitigated by the platform's underlying safety guardrails.

jira-automation