The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides installation instructions for the trylle-cli binary using npm, uv, and cargo. These references target official vendor packages (@trylle/trylle-cli, trylle-cli). The update mechanisms, such as npx skills add trylle-labs/skills and uvx upd-skill, also point to the official repository of the skill author.
[COMMAND_EXECUTION]: The skill uses the try CLI to perform repository operations, secret management, and CI interactions. The documentation includes explicit safety rules, such as discouraging the printing of tokens and recommending the use of environment variables for secret inputs to prevent credential exposure in logs or shell history.
[INDIRECT_PROMPT_INJECTION]: The skill interacts with external content from the Trylle platform (e.g., pull request bodies, issue comments, CI logs), which could contain malicious instructions designed to influence the agent.
Ingestion points: Commands like try pr view, try issue view, and try ci logs ingest content from the remote platform (SKILL.md, references/workflows.md).
Boundary markers: The skill instructs the agent to use the --json flag for structured data parsing and to verify command arguments using --help before execution.
Capability inventory: The skill allows for repository modification, secret updates, and pull request management across its command set.
Sanitization: Instructions advise using --body-file and temporary environment variables for complex or sensitive inputs to maintain separation between instructions and data.

trylle-cli