The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill contains a Dockerfile example that downloads and executes an installation script from Anthropic's official domain (claude.ai). This is a well-known service, and the operation is documented neutrally as a legitimate setup step.
[COMMAND_EXECUTION]: The workflow documentation describes the execution of various CLI tools including Git, npm, and Docker for tasks such as automated commits, PR management, and running tests.
[PROMPT_INJECTION]: The skill describes an indirect prompt injection surface where the agent is instructed to process untrusted data from external sources.
Ingestion points: The agent reads external Pull Request data via gh pr checkout and analyzes local source code using tools like ask-gemini with the @src/ context.
Boundary markers: No specific boundary markers or instructions to ignore embedded commands are provided for these ingestion points.
Capability inventory: The skill uses Bash, Read, and Write tools to execute shell commands and modify the file system based on its analysis.
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from external PRs or codebases before processing.

agentic-workflow