The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of the @alwaysmeticulous/cli tool via npx. It provides instructions for running subcommands like auth, simulate, download, and local to manage visual regression testing workflows.
[EXTERNAL_DOWNLOADS]: The skill describes processes for downloading external artifacts including session recordings, replay archives (screenshots, assets, logs), and web assets from the Meticulous platform to the local machine.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by processing data that can be influenced by external sessions or repository metadata.
Ingestion points: Session titles, descriptions, and replay log content are retrieved and displayed to the agent via the local relevant-sessions and download commands.
Boundary markers: The instructions do not specify the use of delimiters or isolation markers when presenting this externally-sourced session metadata to the agent.
Capability inventory: The environment allows for CLI command execution, filesystem access for storing session data, and interaction with the browser during replays.
Sanitization: There is no mention of sanitization or validation of the metadata strings retrieved from session recordings before they are processed by the agent.
[CREDENTIALS_UNSAFE]: The authentication flow involves storing OAuth tokens in the local data directory. The skill uses the METICULOUS_API_TOKEN environment variable and standard CLI authentication patterns to manage access to the Meticulous API, following industry practices for developer tools.

meticulous-cli