Skip to content

Strengthen lab AGENTS env-development guardrails#876

Merged
willccbb merged 3 commits intomainfrom
codex/update-agents.md-for-env-initialization-guidelines
Feb 9, 2026
Merged

Strengthen lab AGENTS env-development guardrails#876
willccbb merged 3 commits intomainfrom
codex/update-agents.md-for-env-initialization-guidelines

Conversation

@willccbb
Copy link
Member

@willccbb willccbb commented Feb 9, 2026

Motivation

  • Make end-user guidance explicit that environment source code must not be initialized manually and that the Prime CLI is the canonical interface for environment workflows.
  • Prevent inconsistent or messy environment development by requiring prime lab setup before starting env work and by encouraging a dedicated environments/<env_name>/ layout.
  • Ensure agents follow and surface best-practice patterns and redirect users to the supported workflow when requests would deviate from the guidelines.

Description

  • Updated assets/agents/end_user_best_practices.md to add strict instructions: never initialize env source manually, always use prime env init, prefer the Prime CLI for lifecycle operations, require prime lab setup before env development, and keep environments self-contained under environments/<env_name>/.
  • Regenerated the compiled lab guidance by running scripts/sync.py, which updated assets/lab/AGENTS.md to reflect the source changes.
  • Committed the modified source and compiled files with a focused message to make the guardrails discoverable to agents and users.

Testing

  • Ran uv run pre-commit install which completed successfully.
  • Ran uv run python scripts/sync.py to regenerate compiled AGENTS files which completed and updated assets/lab/AGENTS.md.
  • Ran uv run pre-commit run --all-files (includes ruff checks and the sync validation) and it passed.

Codex Task


Note

Low Risk
Documentation-only updates that tighten recommended workflows for lab environment development; no runtime or API behavior changes.

Overview
Strengthens lab end-user guidance to explicitly enforce the Prime CLI as the canonical environment workflow: never manually initialize env source, require prime lab setup before env work, and standardize on a self-contained environments/<env_name>/ layout with documented best-practice patterns.

Regenerates the compiled lab assets/lab/AGENTS.md so it reflects the updated source guidance in assets/agents/end_user_best_practices.md.

Written by Cursor Bugbot for commit 093d114. This will update automatically on new commits. Configure here.

@willccbb willccbb merged commit 3409cb5 into main Feb 9, 2026
3 checks passed
Copy link

@cursor cursor bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Cursor Bugbot has reviewed your changes and found 1 potential issue.

Bugbot Autofix is OFF. To automatically fix reported issues with Cloud Agents, enable Autofix in the Cursor dashboard.

- Document required environment variables in README and validate missing keys early with `vf.ensure_keys(...)`.
- NEVER initialize environment source code manually; ALWAYS create new environments with `prime env init`.
- Use the Prime CLI for all environment lifecycle operations (`prime env init` → `prime env install` → `prime eval run` → `prime env push`) rather than ad-hoc scripts.
- NEVER begin environment development before `prime lab setup` has been run; if work starts outside that structure, recommend adjusting course into a proper lab workspace before continuing.
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Missing skill update for new lab setup prerequisite

Low Severity

The PR introduces a new requirement that prime lab setup must be run before beginning environment development, but skills/create-environments/SKILL.md does not reflect this prerequisite. The skill file starts with prime env init my-env as the first step without mentioning the lab setup requirement. This creates an inconsistency between the agent guidance files and the create-environments skill, potentially causing agents to skip the now-mandatory workspace setup. Per the Skills Updates rule: "If workflow-relevant changes are detected without matching skill updates, request that the author update the impacted skills before merge."

Fix in Cursor Fix in Web

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant