The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the gh issue create command to programmatically interact with the GitHub CLI for creating issues on remote repositories.
[DATA_EXFILTRATION]: By design, the skill extracts information derived from the local codebase (domain language, feature behavior, and behavior boundaries) and transmits it to GitHub. While GitHub is a well-known service, this involves sending project-specific context to an external platform.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted data from both user input and the local codebase files.
Ingestion points: The agent ingests bug descriptions directly from the user and explores various codebase files, including UBIQUITOUS_LANGUAGE.md, to gather context.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to treat the ingested codebase content as data rather than instructions.
Capability inventory: The skill possesses the capability to read any file in the codebase (via the Explore subagent) and write to external repositories (via gh issue create).
Sanitization: The instructions do not define any sanitization, validation, or filtering mechanisms for the data gathered from the user or the codebase before it is included in the GitHub issue body.

qa