The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to execute code from the npm registry using npx designlang and npm install -g designlang. These packages are from third-party sources not included in the trusted vendor list.\n- [REMOTE_CODE_EXECUTION]: The use of npx to run designlang and the command npx skills add Manavarya09/design-extract both download and execute code from remote repositories on the local system.\n- [CREDENTIALS_UNSAFE]: The documentation explicitly provides examples of passing sensitive authentication information, such as session cookies, bearer tokens, and API keys, as plaintext command-line arguments (e.g., --cookie "session=$SESSION_COOKIE", --header "Authorization:Bearer $TOKEN"). This is an insecure practice as it can expose secrets in process lists, shell history, and logs.\n- [COMMAND_EXECUTION]: The skill's primary functionality relies on executing shell commands to crawl websites and manipulate the local filesystem (e.g., apply and clone commands).\n- [INDIRECT_PROMPT_INJECTION]: The skill creates a vulnerability surface for indirect prompt injection by ingesting data from untrusted external websites.\n
Ingestion points: The tool crawls user-specified URLs to extract styles and content (SKILL.md).\n
Boundary markers: The instructions do not specify any boundary markers or warnings to the agent to ignore instructions embedded in the extracted content.\n
Capability inventory: The skill has the ability to write to the filesystem, modify existing projects via the apply command, and scaffold new projects via the clone command (SKILL.md).\n
Sanitization: There is no documentation of sanitization or filtering of the extracted DOM content before it is processed into AI-optimized markdown files.

designlang-design-extract