The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill requires the installation and execution of an external CLI tool (@swiftlysingh/excalidraw-cli) from an unverified third-party source. Instructions provide both npm install and npx execution methods.
[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using npx where user-derived descriptions are interpolated into the command line: npx @swiftlysingh/excalidraw-cli create --inline "YOUR_DSL_HERE". This pattern is susceptible to command injection if the agent generates a DSL string containing shell metacharacters like semicolons, backticks, or command substitutions.
[EXTERNAL_DOWNLOADS]: The skill prompts the user to clone a repository from GitHub (https://github.com/swiftlysingh/excalidraw-skill) and install a package from the NPM registry that is not associated with a known trusted vendor.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted natural language input from the user to generate input for a shell-based tool, creating an injection surface.
Ingestion points: Natural language descriptions of flowcharts provided by the user in SKILL.md (Step 1).
Boundary markers: There are no instructions for the agent to use delimiters or to ignore potential instructions embedded within the user's diagram description.
Capability inventory: The skill utilizes shell execution via npx to run external code.
Sanitization: The instructions lack any requirement for the agent to sanitize or escape the generated DSL string before it is passed to the shell environment.

excalidraw-flowchart