The Agent Skills Directory

[Indirect Prompt Injection]: The skill is designed to ingest and process data from external GitHub pull requests, including titles, descriptions, and community comments.
Ingestion points: The skill captures the PR number, title, URL, body, labels, author, merge time, changed files, diffs, linked issues, and review/author comments (Step 2).
Boundary markers: No explicit delimiters or instructions are provided to the agent to disregard instructions embedded within these external inputs.
Capability inventory: The agent uses the gh CLI to search for and create issues in the microsoft/vscode-docs repository.
Sanitization: There are no specific instructions for escaping or filtering the external content before it is processed or included in new issue descriptions.
Context: Because this is an unattended daily workflow, if a pull request contains text designed to override agent instructions, the agent might inadvertently perform actions contrary to its core task while evaluating documentation impact or creating issues.
[Command Execution]: The skill utilizes the gh CLI to interact with GitHub for data retrieval and issue management.
Evidence: The instructions state: "Use the gh CLI for all GitHub operations."
Context: This is the primary mechanism for the skill's functionality and is restricted to the specific task of managing project documentation within the relevant repositories.

daily-docs-audit