The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill enables an agent to ingest and process data from external Gitea repositories, which may contain attacker-controllable content like issue descriptions or pull request bodies.\n
Ingestion points: tea issues list, tea pulls list, tea notifications list, and tea actions runs logs (found in SKILL.md and CLI_REFERENCE.md).\n
Boundary markers: None provided to distinguish untrusted repository content from instructions.\n
Capability inventory: The skill has extensive capabilities including shell command execution (tea), network communication (Gitea API), and secret management (tea actions secrets create).\n
Sanitization: No sanitization mechanisms are described or required by the instructions.\n- [EXTERNAL_DOWNLOADS]: The documentation provides a method to download the tea CLI binary directly from its official repository on GitHub (https://github.com/gitea/tea/releases). As Gitea is a well-known service, this download is documented as a standard installation procedure.\n- [COMMAND_EXECUTION]: The skill uses a wide array of tea subcommands to perform repository operations, involving local shell execution of the CLI tool. This includes managing authentication, repository settings, and continuous integration secrets.

tea