The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute standard development tools such as git and the GitHub CLI (gh). These tools are used for read-only operations to determine the repository's status, including branch information, unpushed commits, and the state of pull requests and issues. These operations are consistent with the skill's purpose as a workflow sensor.
[PROMPT_INJECTION]: The skill processes untrusted data from project-specific files such as ROADMAP.md, TASKS.md, and known-issues.md, as well as GitHub issue content. While these sources could theoretically contain malicious instructions, the skill mandates a mechanical, read-only approach for data extraction, significantly reducing the risk of the agent executing embedded commands.
Ingestion points: Local documentation files (docs/features/ROADMAP.md, docs/fix/README.md) and feature-specific task/progress files; GitHub issue and PR metadata retrieved via the gh tool.
Boundary markers: Absent; the skill does not explicitly instruct the agent to ignore prompt-like content within the external files it reads.
Capability inventory: Read-only access to the local file system and repository metadata via git and gh. No write permissions, arbitrary code execution, or unauthorized network access are defined.
Sanitization: Absent; the content from documentation files is used to populate a structured JSON envelope without specific escaping or filtering.

workflow-status