The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automatically detects and executes build and verification tools present in the repository, such as npm test, cargo test, and ./gradlew build. This involves executing local scripts and binaries within the project environment based on instructions derived from external data.
[INDIRECT_PROMPT_INJECTION]: The skill fetches untrusted data from external GitHub issues and comments and uses this content as the primary source for planning and implementing code modifications.
Ingestion points: Data enters the agent's context through the gh issue view command, specifically the body and comments fields of the issue.
Boundary markers: Absent. The skill instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when the agent processes the fetched content.
Capability inventory: The agent has access to Bash, Edit, and Write tools, providing the ability to execute arbitrary shell commands and modify any file in the repository.
Sanitization: Absent. There is no evidence of filtering, validation, or escaping of the issue content before it is interpolated into the agent's planning and implementation process.

github-do-issue