The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill directs the agent to install the judge-me-pp-cli tool using npx from the @mvanhorn/printing-press-library NPM package and go install from the github.com/mvanhorn/printing-press-library repository. These resources are provided by the tool's vendor.
[COMMAND_EXECUTION]: The skill's primary functionality is achieved by executing shell commands via the judge-me-pp-cli binary to interact with the Judge.me REST API and local SQLite storage.
[DATA_EXFILTRATION]: The CLI tool includes a --deliver webhook:<url> feature that allows routing command output to a user-specified URL via HTTP POST. While a standard feature for automation, it represents a potential data egress point.
[PROMPT_INJECTION]: The skill processes untrusted review data from external sources. The documentation explicitly warns that raw review information is not sanitized, creating an attack surface for indirect prompt injection or cross-site scripting (XSS) if the output is rendered without further processing.
Ingestion points: reviews index, reviews get, and reviews update commands in SKILL.md fetch content from the Judge.me API.
Boundary markers: No specific delimiters or "ignore" instructions are implemented for the interpolation of fetched review content into the agent's context.
Capability inventory: The agent has access to shell execution (bash) and network operations through the CLI tool.
Sanitization: The skill contains an explicit warning that raw review data is not sanitized.

pp-judge-me