The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run to execute external binaries including acli (Atlassian CLI), gog (Google Workspace CLI), and git. These are used for legitimate data gathering from Jira and Google Sheets as part of the skill's primary purpose. Commands are correctly passed as lists to the subprocess call, which is a best practice for preventing shell injection.
[DYNAMIC_EXECUTION]: scripts/release.py dynamically modifies the Python search path (sys.path) to import a shared module (adf_milestones) from a sibling skill (rhdh-jira-api). This represents a vendor-internal dependency for sharing common Jira field extraction logic across related skills.
[INDIRECT_PROMPT_INJECTION]: The skill represents a potential vulnerability surface for indirect prompt injection because it ingests untrusted data from external sources (Jira issue descriptions and Google Sheets rows) which are then processed by the agent.
Ingestion points: scripts/release.py (via Jira search results and Google Sheets rows) and scripts/rich_filter.py (via Rich Filter JSON exports).
Boundary markers: None identified in the prompt templates.
Capability inventory: subprocess.run calls for Jira, Google Sheets, and Git operations are present across multiple scripts.
Sanitization: No specific sanitization or filtering of the external data before presentation to the agent was observed.

rhdh-release-schedule