The Agent Skills Directory

[SAFE]: The skill uses environment variables (CONFLUENCE_API_TOKEN) or a local MCP configuration file for authentication, avoiding hardcoded secrets and following standard secret management practices.
[SAFE]: The Python script scripts/confluence_adf.py uses only the standard library, eliminating risks associated with third-party package supply chains.
[DATA_EXFILTRATION]: The script includes a custom NoRedirect HTTP handler designed to prevent urllib from forwarding sensitive Basic Authentication headers if the Confluence site issues a redirect to a different domain.
[SAFE]: API interactions are strictly validated. The Client.call method ensures that constructed URLs remain within the configured site context, and pagination link processing includes hostname and scheme validation to prevent SSRF or redirection to untrusted targets.
[COMMAND_EXECUTION]: The instructions for the agent involve executing the local script with specific parameters. No patterns of arbitrary command injection or unsafe string interpolation into shell commands were found.

confluence-adf