The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructs the agent to run npx -y @nomadamas/k-skill@0 as a "required first step" to obtain instructions. This command downloads and executes code from an unverified NPM package scope, which is not associated with any trusted vendor or the skill author's verified infrastructure.
[EXTERNAL_DOWNLOADS]: The skill downloads code from the NPM registry and makes network requests to k-skill-proxy.nomadamas.org. It also references remote files at github.com/NomaDamas/k-skill to define its core behavior.
[COMMAND_EXECUTION]: The skill relies on shell commands including npx (for package execution) and curl (for API requests) to perform its functions, including fetching the primary instruction set.
[INDIRECT_PROMPT_INJECTION]: The skill functions as a stub that pulls instructions and data from remote sources at runtime, creating a significant attack surface for indirect prompt injection.
Ingestion points: SKILL.md directs the agent to fetch instructions via npx and instruction.md pulls data from k-skill-proxy.nomadamas.org.
Boundary markers: No boundary markers or delimiters are used to wrap or sanitize the remote content.
Capability inventory: The skill has access to shell execution (npx, curl).
Sanitization: There is no evidence of sanitization or validation performed on the instructions or data retrieved from remote sources.

han-river-water-level