The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches visual design guidelines from Vercel's official documentation (https://vercel.com/design.md). This is a trusted source for styling information.
[REMOTE_CODE_EXECUTION]: The skill uses npx to download and execute the lavish-axi package to generate and review plans. While it defaults to the author's GitHub repository (github:Timmyy3000/lavish-axi), the package identifier is resolved from the LAVISH_AXI_PACKAGE variable, which can be overridden to point to arbitrary remote code.
[COMMAND_EXECUTION]: Executes shell commands via npx to manage Lavish editor sessions, including design, open, poll, end, and export commands.
[DYNAMIC_EXECUTION]: Dynamically constructs shell commands by substituting the <lavish-package> placeholder with a resolved variable (LAVISH_AXI_PACKAGE) before execution.
[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted content from the target repository to generate plans, creating an attack surface for indirect prompt injection.
Ingestion points: Reads repository instructions (AGENTS.md, .cursor/rules/), plan templates, and backend/frontend conventions (SKILL.md, references/delegation.md).
Boundary markers: None identified; the skill does not use specific delimiters or instructions to separate ingested content from its own logic.
Capability inventory: Includes file writing, npx command execution, and network fetching (SKILL.md).
Sanitization: No explicit sanitization or validation of the ingested repository content is performed before processing.

plan-it