The Agent Skills Directory

[COMMAND_EXECUTION]: The skill supports the execution of arbitrary local commands via the on_answer hook defined in the config.yaml file, which runs whenever a user provides an answer on the local board.
[COMMAND_EXECUTION]: The post --type signin command supports a --start argument that executes a shell command (e.g., npm login) on the host machine to generate sign-in links.
[REMOTE_CODE_EXECUTION]: The documentation instructs the agent to install and execute a third-party skill from an unknown author using npx skills add tt-a1i/archify.
[REMOTE_CODE_EXECUTION]: The skill relies on npx -y sign-off-axi to download and execute its primary CLI tool from the npm registry at runtime.
[DYNAMIC_EXECUTION]: The skill utilizes a 'standalone owner dispatcher' mechanism to manage and resume agent sessions, which involves local process management and command-based acknowledgments.
[INDIRECT_PROMPT_INJECTION]: The skill has a high attack surface for indirect injection as it is designed to ingest and act upon data (answers, notes, and comments) provided by users through external local JSON files.
Ingestion points: The wait and answers commands read data from .sign-off/ board files.
Capability inventory: The skill can execute commands, read/write local files, and manage network-based sign-in flows.
Boundary markers: The documentation does not specify the use of delimiters or sanitization for user-provided text before it is processed by the agent.
Sanitization: No explicit sanitization or validation of the contents of the answer files is described.

sign-off-axi