Install AI Guardrails

Install lightweight AI coding guardrails in a repository without changing product behavior or scanning generated files.

CopilotCursorKilo
AuthorNeexoCore
Updated
guardrailsonboarding

You are installing lightweight AI coding guardrails into this repository.

Do not implement product features. Do not refactor application code. Do not scan the whole repository.

Inspect only:

  • README or equivalent project overview
  • package metadata
  • top-level folder structure
  • existing .github folder
  • obvious config files
  • existing docs related to architecture, testing, deployment, or AI usage

Create or update only when relevant:

  • .github/copilot-instructions.md
  • .github/instructions/*.instructions.md
  • .github/prompts/*.prompt.md
  • docs/AI_CONTEXT.md
  • docs/AI_USAGE_POLICY.md
  • docs/COPILOT_CONTENT_EXCLUSIONS.md
  • docs/GITHUB_ADMIN_CHECKLIST.md
  • docs/MODEL_USAGE_GUIDE.md
  • AGENTS.md

Rules:

  • keep always-on instruction files short
  • do not duplicate the same rule across every file
  • do not read generated files, build output, binary assets, lock files, logs, or large media/3D assets
  • if uncertain, write Unknown - verify manually
  • do not change application behavior
  • do not add dependencies

Final response must include:

  1. files changed
  2. project-specific assumptions
  3. recommended GitHub content exclusions
  4. manual GitHub settings still required
  5. risks

Raw content

Copy into your project — e.g. .instructions.md, .agent.md, or SKILL.md

You are installing lightweight AI coding guardrails into this repository.

Do not implement product features.
Do not refactor application code.
Do not scan the whole repository.

Inspect only:

- README or equivalent project overview
- package metadata
- top-level folder structure
- existing `.github` folder
- obvious config files
- existing docs related to architecture, testing, deployment, or AI usage

Create or update only when relevant:

- `.github/copilot-instructions.md`
- `.github/instructions/*.instructions.md`
- `.github/prompts/*.prompt.md`
- `docs/AI_CONTEXT.md`
- `docs/AI_USAGE_POLICY.md`
- `docs/COPILOT_CONTENT_EXCLUSIONS.md`
- `docs/GITHUB_ADMIN_CHECKLIST.md`
- `docs/MODEL_USAGE_GUIDE.md`
- `AGENTS.md`

Rules:

- keep always-on instruction files short
- do not duplicate the same rule across every file
- do not read generated files, build output, binary assets, lock files, logs, or large media/3D assets
- if uncertain, write `Unknown - verify manually`
- do not change application behavior
- do not add dependencies

Final response must include:

1. files changed
2. project-specific assumptions
3. recommended GitHub content exclusions
4. manual GitHub settings still required
5. risks

Next steps

Matched by shared tags and category. Browse all cookbook