The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection because it ingests untrusted data from external GitHub issues without sufficient boundary markers or sanitization.
Ingestion points: The workflow begins by fetching issue content using gh issue view <number>, which imports the issue title, body, and comments directly into the agent's context.
Boundary markers: The instructions do not define delimiters or provide explicit 'ignore embedded instructions' warnings when processing the issue content.
Capability inventory: The skill possesses significant capabilities, including filesystem writes, command execution via bun and git, and the ability to open pull requests using gh.
Sanitization: No validation or sanitization of the fetched issue body is performed before the agent uses it to plan and implement the fix.
[COMMAND_EXECUTION]: The skill executes several shell commands to manage the development workflow. This includes git for branch management, gh for GitHub interaction, and bun for build and dependency management. While these are necessary for the skill's functionality, they represent a powerful capability set that could be abused if the agent is influenced by malicious issue content.
[EXTERNAL_DOWNLOADS]: The skill performs external operations such as fetching data from GitHub (git fetch, gh issue view) and downloading packages via bun install. These operations target well-known services and registries associated with the software development lifecycle.

issue-fixer