The Agent Skills Directory

[SAFE]: The skill instructions and workflows strictly adhere to standard Atlassian administration procedures and security best practices, using official API endpoints and management portals.
[SAFE]: The scripts/permission_audit_tool.py script performs static analysis on local JSON data using standard libraries (argparse, json, sys). It does not use network modules or perform any dynamic code execution.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it is designed to process data from Atlassian APIs (e.g., Jira issue descriptions, user properties) and local JSON files to perform sensitive administrative tasks. However, this is inherent to its primary purpose as an administration tool.
Ingestion points: The permission_audit_tool.py script reads a JSON permissions file; workflows in SKILL.md process responses from Atlassian REST APIs.
Boundary markers: Absent; the skill does not define specific delimiters or instructions to distinguish between external data and system commands.
Capability inventory: The agent has the authority to delete users, modify group memberships, change permission schemes, and install marketplace apps.
Sanitization: Not explicitly present in the provided scripts or workflow instructions.

atlassian-admin