The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses destructive file system commands, specifically rm -rf, to delete framework directories during the project cleanup process described in references/framework-cleanup.md. While this is for an intended cleanup purpose, recursive deletion of directories is a high-risk operation.
[COMMAND_EXECUTION]: Automated database management tasks are performed using pg_isready for environment detection and createdb for database instantiation. These scripts include shell logic that attempts execution under different user contexts (e.g., postgres, whoami).
[CREDENTIALS_UNSAFE]: The skill contains hardcoded default credentials intended for initial seed data and local testing, specifically admin123 and user123456 in SKILL.md. It also specifies a hardcoded password (tinyship123) for containerized PostgreSQL instances.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download and execution of external software by pulling the postgres:15 official image from Docker Hub and installing project dependencies using pnpm install from the standard registry.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process several local project files to guide the setup wizard, creating a surface for potential instruction injection if these files are externally controlled.
Ingestion points: env.example, pnpm-workspace.yaml, package.json, turbo.json, docker-compose.yml, .github/workflows/ci.yml, AGENTS.md, and project documentation files.
Boundary markers: Absent. The skill lacks explicit delimiters or instructions to ignore embedded commands within the processed files.
Capability inventory: Recursive directory deletion (rm -rf), database creation (createdb), container execution (docker run), package installation (pnpm install), and file system manipulation (cp, mkdir).
Sanitization: No evidence of content sanitization or validation was found for the data read from the local files before it is processed by the agent.

tinyship-setup