Menu
Tangram command /tangram:conditioning. Run automated QA, including type checking, linters, and test suites to verify project stability.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
AUTOMATIC TRIGGER: Activate this skill immediately when the user asks to create, scaffold, initialize, or build an Astro project, a Cloudflare Worker, or a full-stack app on Cloudflare. The user does NOT need to explicitly name this skill or use a slash command. Use this autonomously for ANY request related to starting an Astro JS project. Scaffolds Astro + Cloudflare Workers, Durable Objects, Elysia API, D1, and DDD architecture.
Deep-dive into the system architecture, flow, and structural logic using modern patterns and internet research. Codex repo skill for the legacy /tangram:design-architecture workflow.
Deep-dive into hosting, CI/CD pipelines, and environment management. Codex repo skill for the legacy /tangram:design-deployment workflow.
Deep-dive into authentication, authorization, and data protection strategies. Codex repo skill for the legacy /tangram:design-security workflow.
Deep-dive into the technology stack, selecting the best tools and versions based on current ecosystem trends. Codex repo skill for the legacy /tangram:design-stack workflow.
Deep-dive into the file and folder hierarchy based on the selected tech stack and architecture. Codex repo skill for the legacy /tangram:design-structure workflow.
| name | tangram-conditioning |
| description | Tangram command /tangram:conditioning. Run automated QA, including type checking, linters, and test suites to verify project stability. |
Codex adaptation of .codex/workflows/tangram/commands/tangram/maintain/conditioning.toml.
Use this skill when the user asks for /tangram:conditioning, $tangram-conditioning, or the corresponding Tangram workflow in natural language. Codex does not load source workflow .toml command files directly; this SKILL.md carries the converted prompt.
You are the Tangram Build AI executing the conditioning command. Your goal is to run the project's health checks to ensure stability.
Input: Triggered by /tangram:conditioning.
Hierarchy of Truth
Steps
Stack Identification Read tangram/design/stack.md and package.json (or equivalent) to determine the testing and linting tools available (e.g., ESLint, Prettier, PyTest, Jest, TSC).
The Conditioning Sequence Execute the health checks in terminal (using appropriate shell tools):
Log the Results Capture the stdout/stderr from these commands. Identify any warnings or hard failures.
Suggest Action
Output On Success
Conditioning Complete
Checks Run: [Lint, Types, Tests] Results: [Pass / X Errors Found]
Log:
- Linting: Passed
- Type Check: Passed
- Tests: Passed X / Failed Y
Status: Stable / Requires Debugging
Guardrails
