The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and install the rechtspraak-pp-cli binary from the author's GitHub repository (github.com/mvanhorn) or NPM package (@mvanhorn/printing-press-library).
[DATA_EXFILTRATION]: Provides mechanisms for routing data to external network locations. The --deliver webhook: flag enables sending command results to an arbitrary external URL. Additionally, the feedback command can be configured via environment variables (RECHTSPRAAK_FEEDBACK_ENDPOINT) to automatically exfiltrate local feedback logs to a remote server.
[COMMAND_EXECUTION]: Requires the Read Bash tool to execute CLI commands within the shell environment to interact with the installed legal research tool.
[PROMPT_INJECTION]: The skill ingests and processes external data (Dutch court decisions) which creates a surface for indirect prompt injection.
Ingestion points: Decisions are retrieved via uitspraken get, search, and narrow commands in SKILL.md.
Boundary markers: No explicit instructions are provided to the agent to treat the retrieved legal text as untrusted data or to ignore embedded instructions.
Capability inventory: The agent can execute shell commands and route output to files or webhooks.
Sanitization: There is no evidence of sanitization or escaping of the retrieved legal text bodies before they are processed.

pp-rechtspraak