The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install the 'mex-agent' package globally via NPM. This downloads executable code from the public NPM registry to the local system.
[COMMAND_EXECUTION]: The skill includes a bash script at scripts/mex.sh that acts as a wrapper for the mex CLI. It executes commands like mex check and mex update with arguments passed from the agent. It also uses allowed-tools such as Bash and Edit to modify project configuration files like .mex/ROUTER.md.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of reading and processing project files.
Ingestion points: The skill reads project source files (TypeScript, JavaScript, Python, Rust) and user-generated markdown files within the .mex/wiki/ directory to build architectural context.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following malicious instructions that might be embedded in code comments or wiki pages.
Capability inventory: The skill has broad capabilities including file system access (Read, Write, Edit, Glob) and shell command execution (Bash).
Sanitization: No sanitization or filtering mechanisms are described for the content extracted from the project files before it is provided to the agent as context.

mex