The Agent Skills Directory

[COMMAND_EXECUTION]: The skill operates primarily through shell command execution using the try CLI and git for version control and platform management.
[EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the try CLI via Homebrew using brew install trylle/trylle-cli/try. This targets a vendor-managed repository necessary for the skill's operation.
[CREDENTIALS_UNSAFE]: The skill documents commands that access sensitive credentials, such as try auth token and try bot token. However, it includes explicit safety rules advising against routine inspection of tokens and recommending the use of environment variables and cleanup steps (unset) for secret management.
[PROMPT_INJECTION]: The skill processes untrusted external data, which constitutes an indirect prompt injection surface.
Ingestion points: The skill reads data from repository content, pull request diffs, issue comments, and CI logs (e.g., via try pr view, try issue view, try ci logs).
Boundary markers: There are no explicit instructions to wrap or delimit ingested external content to prevent the agent from obeying embedded instructions.
Capability inventory: The agent has access to powerful write operations, including merging pull requests, deleting repositories, and managing organization secrets.
Sanitization: No specific sanitization or filtering of external content is defined in the instructions.

tylle-cli