Advanced Standards-Intelligence Criteria

Compatibility note: this file keeps the historical docs/v5-scope.md path and V5-linked internal identifiers for scripts, tools, and tests. Public product language should describe this as an advanced standards-intelligence maturity layer, not as a product version.

This criteria layer remains local-first. It excludes hosted service work, database/storage work, GitHub/PR adapters, CLI UX, and typed SDKs.

It should make architect-mcp more self-improving, more explainable, and easier for agents to apply correctly inside the current MCP tool surface.

Standards-Intelligence Direction

This layer focuses on local intelligence and usability of the standards harness:

• Standards orchestration.
• Explainable findings.
• Policy simulation.
• Cross-stack conflict handling.
• Eval expansion.
• Agent teaching and handoff quality.
• Repo-profile intelligence.
• Contract lifecycle quality.

Standards Orchestration

Goal: make standards selection smarter without adding storage or network dependencies.

Deliverables:

• A standards resolver that explains why each foundation, policy, and stack pack was selected.
• Conflict ranking when multiple packs apply to the same files or concerns.
• A "minimum viable standards" mode for small projects.
• A "strict production standards" mode for high-risk projects.
• A standards coverage report by project brief, stack, and file summary signals.

Explainable Findings

Goal: make review findings easier for novice users and agents to act on.

Deliverables:

• Plain-English finding explanations.
• "What this probably means" sections for common terms like auth, API boundary, migration, dependency risk, and accessibility.
• Suggested next tool calls for each finding.
• Fix-shape recommendations: split file, add boundary, add test, ask user, verify, or defer.
• Evidence snippets showing which file signals triggered each finding.

Policy Simulation

Goal: let clients preview how stricter or looser policy settings would affect a repo before enforcing them.

Deliverables:

• Simulate review gates across summary, migration, ci, and custom thresholds.
• Show which findings would become blockers under stricter modes.
• Show which baselines or accepted findings would stop suppressing issues.
• Preview impact of adding a new policy pack or stack pack.

Cross-Stack Conflict Handling

Goal: improve behavior when multiple stack packs give overlapping or competing advice.

Deliverables:

• Better conflict explanations.
• Conflict severity ranking.
• Suggested resolution: prefer framework pack, prefer policy pack, ask user, or split path scopes.
• Tests for overlapping auth, validation, frontend, backend, and AI/tool rules.

Eval Expansion

Goal: make the harness harder to regress without relying on live services.

Deliverables:

• Standards-intelligence eval suites for novice-intent interpretation, standards selection, explainable findings, policy simulation, conflict resolution, and handoff quality.
• Golden fixture reports for common project archetypes.
• Regression fixtures for known agent failure modes.
• Eval summaries that explain failures in product language, not just assertion messages.

Agent Teaching

Goal: help agents and novice users understand what to do next without turning findings into noisy lectures.

Deliverables:

• Concise teaching snippets attached to high-value findings.
• Terminology mapping for common novice phrases.
• One-question clarification suggestions for ambiguous fixes.
• "Do this / avoid this / prove it with this" output blocks.
• Agent handoff summaries that are short enough to reuse in the next turn.

Repo Profile Intelligence

Goal: make existing repo layout handling stronger without adding external integrations.

Deliverables:

• More built-in repo profiles.
• Profile confidence scoring from file summaries.
• Profile mismatch warnings when a contract forces the wrong structure.
• Suggested repoLayout.pathMap updates.
• Fixtures for monorepos, packages/apps layouts, mobile apps, docs-heavy repos, and mixed frontend/backend repos.

Contract Lifecycle

Goal: make architecture contracts easier to evolve safely.

Deliverables:

• Contract maturity levels.
• Contract diff explanations in plain English.
• Safer rule deprecation and replacement guidance.
• Contract changelog generation from diffs.
• Better warnings when new rules would create too many findings for an existing repo.

Non-Goals

• Hosted service or hosted API implementation.
• Database/storage, including schema, persistence, migrations, review history, or memory storage.
• GitHub repository, branch, PR, issue, app, OAuth, or comment workflow.
• CLI UX.
• Typed client SDK.
• Billing, accounts, teams, or dashboards.

Those remain out of scope until a later backend/productization phase.