Build Right Engineering Principles

Build Right Engineering Principles

Use this skill as a cross-cutting engineering standard for Build Right projects. It is not a lifecycle phase. It is a review and design lens that other Build Right skills may load when architecture, contracts, implementation boundaries, or engineering-quality tradeoffs matter.

Required Reading

• Read references/principles.md before making or reviewing changes that touch

architecture boundaries, public contracts, provider adapters, generated code, package ownership, side effects, errors, observability, security, testing strategy, or enforceable policy.

• Do not load the reference for routine product capture, backlog grooming, or

narrow content-only updates unless those topics raise engineering-risk questions.

Operating Mode

1. Identify the changed or proposed engineering surface:

architecture, module responsibility, dependency direction, contract, adapter, state shape, side effect, failure path, test, evidence, or policy.

1. Apply the review checklist from references/principles.md.
2. Separate guidance from enforceable authority. Markdown principles are

guidance until backed by code, checks, tests, schemas, policy, or evidence.

1. Prefer the smallest correction that preserves local patterns and reduces

real risk. Do not introduce abstraction only because a principle names one.

1. When repeated findings expose a process gap, recommend an enforcement

surface instead of adding more prose.

1. Record significant architecture or public-contract choices in the target

repo's normal decision surface.

Closeout

End with the engineering result that matters:

Principles applied: 
Required changes: 
Enforcement gap: 
Residual risk: 

User-Visible Status Badge

End every final response with exactly one status badge block:

✅ [DONE] Status: DONE
Decision: 
Next action: 
Needs user input: 
Blocked by: 

Use this status map:

• ✅ [DONE] Status: DONE when there are no required changes and

no enforcement gap.

• 🟢 [GREEN] Status: ALL GREEN when the reviewed direction is safe to

continue but implementation is not complete.

• 🟡 [YELLOW] Status: NEEDS INPUT when a user or owner architecture,

contract, security, or policy decision is needed.

• 🟠 [ORANGE] Status: NEEDS WORK when AI-owned corrections, tests,

docs, or enforcement work remain.

• 🔵 [BLUE] Status: WAITING EXTERNAL for external proof, credentials,

production access, publishing, indexing, or third-party state.

• 🔴 [RED] Status: BLOCKED for unresolved conflicts, failed

validation, source mismatch, invalid state, or enforcement blockers.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: pax-k
• Source: pax-k/build-right
• License: MIT

Install and usage instructions live in the source repository linked above.