The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill's suggested configuration uses npx -y @sooperset/mcp-atlassian, which automatically downloads and executes code from an unverified NPM package scope at runtime.- [EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch the MCP server implementation from the @sooperset NPM registry scope and refers to external documentation hosted on GitHub Pages.- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its core functionality of ingesting and acting upon third-party data. Ingestion points: Reads Jira issue fields, comments, and Confluence page content (SKILL.md). Boundary markers: The instructions lack explicit delimiters or specific 'ignore embedded instructions' directives when processing retrieved Atlassian data. Capability inventory: The skill possesses write capabilities including creating tickets, updating tickets, and editing wiki pages (SKILL.md). Sanitization: Although the skill includes a 'MUST NOT' constraint regarding skipping sanitization, it provides no technical implementation or reinforced prompt constraints to ensure the agent ignores malicious instructions found within the data it reads.

atlassian-mcp