The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill facilitates the execution of the programmable-secret utility via npx. This pattern involves downloading and executing remote code from the NPM registry at runtime. As the package is a verified resource from the skill's vendor, this is considered standard functionality.
[COMMAND_EXECUTION]: The skill instructions include commands for local environment configuration and credential management. Specifically, the env-bootstrap command is used to generate environment files for storing sensitive Ethereum private keys used in blockchain transactions.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface through the ingestion of external data in command arguments, such as metadata JSON and agent URIs used in registration workflows.
Ingestion points: External data is ingested via --metadata-json and --agent-uri flags in the datasets register and identity register commands within SKILL.md.
Boundary markers: No specific delimiters or instruction-isolation warnings are present in the provided command templates.
Capability inventory: The associated CLI tool possesses capabilities for network interaction and local filesystem writes.
Sanitization: The skill documentation does not explicitly detail input validation or sanitization for the provided JSON metadata or URI strings.

programmable-secrets