The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the pegasus CLI to perform project management tasks, such as pegasus projects list, create, update, and show. These commands are used to interact with the SaaS Pegasus API for project configuration.
[SAFE]: The skill checks for existing credentials in standard locations such as ~/.pegasus/credentials or the PEGASUS_API_KEY environment variable. This is documented as the legitimate authentication method for the Pegasus CLI and does not involve exfiltration to unknown domains.
[PROMPT_INJECTION]: The skill includes a surface for indirect prompt injection via the ingestion of configuration files.
Ingestion points: The agent can load configuration data from local YAML or JSON files using the --config-file flag as described in the Commands section of SKILL.md.
Boundary markers: Instructions explicitly mandate the use of the --json flag for CLI output to ensure the agent processes structured data rather than potentially truncated or formatted human-readable text.
Capability inventory: The skill is restricted to project configuration management commands within the pegasus CLI, including creating and updating project settings.
Sanitization: Input validation for fields like project_slug and choice-type fields is handled by the server-side API and the CLI tool, which rejects invalid identifiers or unauthorized feature selections based on the user's license tier.

pegasus-projects