The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/list_azdo_repos.py executes the az and git command-line tools using subprocess.run to list projects, fetch metadata, and identify local clones. All subprocess calls use list-based arguments, which effectively prevents command injection vulnerabilities.- [DATA_EXFILTRATION]: The skill performs network operations by calling the az CLI, which communicates with the well-known service dev.azure.com. It reads and writes project descriptions to Azure DevOps as part of its primary functionality.- [INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection as it processes repository descriptions from Azure DevOps and includes them in a generated ALL_REPOS.md file for the agent to read.
Ingestion points: scripts/list_azdo_repos.py fetches external project descriptions from Azure DevOps.
Boundary markers: The output is formatted as a Markdown table in ALL_REPOS.md.
Capability inventory: The skill can perform file system operations and execute az and git commands.
Sanitization: The script does not sanitize the contents of the descriptions before writing them to the summary file.

cross-repo-discovery