The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill provides instructions for starting the PostgreSQL service using elevated permissions on Linux systems via the sudo service postgresql start command.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read and process content from untrusted project files to determine the execution environment.
Ingestion points: package.json, requirements.txt, and TECH_STACK.md (specified in the Entry Router section).
Boundary markers: Absent; the agent is directed to read these files directly to determine its next set of instructions.
Capability inventory: The skill facilitates extensive shell command execution across various runtimes (Node.js, Python, Go, Rust, Ruby, PHP), database management, and process termination (kill -9).
Sanitization: No sanitization or validation of the input file content is defined before the agent performs routing based on the detected framework.
[EXTERNAL_DOWNLOADS]: The skill references the installation of external development utilities from public repositories, such as air for Go (github.com/cosmtrek/air) and cargo-watch for Rust, using their respective package managers.
[COMMAND_EXECUTION]: The skill facilitates a wide range of terminal commands for local development, including dependency installation (npm install, pip install), environment setup (cp .env.example .env.local), and process management (kill -9 [PID], taskkill /PID [PID] /F).

vibe-coding-local-runner