The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructions in SKILL.md and instruction.md direct the agent to use npx -y @nomadamas/k-skill@0, which downloads and executes the project's CLI tool from the NPM registry. This is a standard mechanism for this skill's ecosystem but involves third-party code execution.
[COMMAND_EXECUTION]: The skill relies on shell command execution via npx to perform searches (slang_search.py), lookups (slang_lookup.py), and to retrieve updated instructions.
[INDIRECT_PROMPT_INJECTION]: The skill fetches and summarizes content from an external source (namu.wiki). This creates a surface for indirect prompt injection if the wiki content contains adversarial instructions designed to influence the agent's behavior during the writing process.
Ingestion points: scripts/slang_lookup.py fetches HTML content from namu.wiki based on user queries or term URLs.
Boundary markers: The skill does not explicitly use boundary markers or delimiters when presenting the fetched summary to the agent, though it does strip HTML tags.
Capability inventory: The skill has network read access via urllib (restricted to namu.wiki and en.namu.wiki) and local file read access for JSON indices.
Sanitization: scripts/slang_lookup.py uses several regular expressions to strip HTML tags, scripts, styles, and navigation elements from the fetched content before returning the text summary.

korean-slang-writing