The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands like npx tsc, npm run, and npx vitest for project diagnostics, type-checking, and test validation. This is standard and expected behavior for a developer-oriented skill.
[DYNAMIC_EXECUTION]: Employs node -e to execute small JavaScript snippets that parse the local package.json file. This is used to detect project dependencies and configuration (e.g., identifying if the project uses Vitest or Biome) and represents a low-risk, common pattern.
[INDIRECT_PROMPT_INJECTION]: The skill has a potential surface for indirect prompt injection because it ingests untrusted project files which could contain malicious instructions.
Ingestion points: Reads project-specific files such as package.json, tsconfig.json, and source code via Read, Grep, and Glob tools.
Boundary markers: The instructions do not define specific delimiters or warnings for the agent to ignore instructions that might be embedded within the project files.
Capability inventory: The agent can execute shell commands, including package manager scripts and testing tools, which could be triggered by data found in ingested files.
Sanitization: No explicit sanitization or filtering of the ingested file content is performed prior to the agent processing the data.

typescript-expert