The Agent Skills Directory

[SAFE]: No malicious patterns, prompt injection, or obfuscation were detected. The skill's operations are transparent and align with its stated purpose of assisting with TypeScript development.
[COMMAND_EXECUTION]: The instruction set includes a 'Memory Protocol' which directs the agent to execute a local cat command on .claude/context/memory/learnings.md. This is a common pattern for maintaining cross-session state in specific agent frameworks.
[DATA_EXPOSURE]: The Node.js utility script (scripts/main.cjs) reads local project metadata (package.json, tsconfig.json) and source files in the src/ directory to report on configuration strictness and use of the any type. This access is read-only and restricted to the workspace.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data (source code) and possesses significant capabilities (Bash, Write, Edit). However, the provided analysis script uses non-executing regex matching to scan content, and the risk is considered negligible for a developer tool of this nature.

typescript-expert