The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The documentation includes a hardcoded PostgreSQL connection string postgresql://postgres:simstudio_local_2026@localhost:5433/simstudio which contains a plaintext password.
[COMMAND_EXECUTION]: The skill provides shell commands for managing Docker services (up, down, logs, restart) and performing administrative API requests using curl against a local endpoint.
[EXTERNAL_DOWNLOADS]: Instructions recommend using npx -y mcp-remote to run the MCP proxy, which downloads and executes an unversioned package from the npm registry at runtime.
[PROMPT_INJECTION]: The skill describes patterns for research and data processing that ingest untrusted external content, creating a surface for indirect prompt injection. • Ingestion points: External content retrieved via brave-search and playwright tools. • Boundary markers: None; the skill does not specify delimiters or instructions for the agent to ignore embedded commands in fetched data. • Capability inventory: Local command execution (Docker, curl), tool orchestration, and database interaction. • Sanitization: No validation or sanitization of external content is mentioned in the workflow patterns.

sim-studio