The Agent Skills Directory

Dynamic Execution (MEDIUM): The file scripts/tests/test_markdown_to_storage.py uses importlib.util to dynamically load and execute confluence_cli.py.
Evidence: spec.loader.exec_module(module) in scripts/tests/test_markdown_to_storage.py executes code from a path calculated at runtime.
Indirect Prompt Injection (LOW): The skill ingests untrusted data from external Confluence pages which could contain malicious instructions for the agent.
Ingestion points: get_page, get_page_by_title, and search_cql in scripts/confluence_api_client.py fetch external content.
Boundary markers: None detected in the provided scripts or instructions.
Capability inventory: The skill can write data back to Confluence via update_page, create_page, and attach_file.
Sanitization: No explicit sanitization or filtering of the fetched content is performed before it is returned to the agent.
Credential Handling (SAFE): The ConfluenceConfig class in scripts/confluence_api_client.py handles API tokens. While no secrets are hardcoded, the script is designed to process sensitive credentials provided by the user.

confluence-cli