Screen 1 — New Audit
Repository input
- GitHub repository URL (public)
- ZIP file upload
- Local path (internal use)
No GitHub OAuth required for MVP. OAuth and GitHub App come later.
Understand which architectural decisions
can actually be governed.
Give Mneme a repository. It identifies architectural decisions, classifies their governability, and shows you exactly which ones translate to deterministic enforcement rules.
Three screens. One actionable audit.
Screen 2 — Audit Overview
Decisions summary
- Total decisions identified
- Enforceable / Partial / Guidance counts
- Coverage percentage
- Sources found (ADRs, CLAUDE.md, etc.)
- Clickable decision list with governability badges
Screen 3 — Decision Detail
Source + rationale + rule
- Full requirement text
- Source file and line numbers
- Governability classification with reasoning
- Applies-to paths
- Proposed Mneme rule (copyable)
What the audit tells you
Decisions discovered
ADRs, CLAUDE.md, AGENTS.md, architecture docs, config files — every source of architectural intent in the repository.
Governability classified
Each decision rated: Enforceable, Partial, or Guidance only.
Proposed Mneme rules
See the exact deterministic Mneme rule where one exists (e.g., FORBID_LITERAL).
Governance gaps
Decisions that can't be enforced yet — with specific next steps to make them machine-testable.
Exportable report
Download the full audit as Markdown or JSON to share with your team or design partners.
Designed for design partners
This audit is the artifact you send after a design-partner call — concrete evidence of where their architecture is ambiguous to AI agents.
The important distinction
Don't say:
"Mneme found architectural drift."
The audit asks:
"Can this architectural intent be deterministically governed?"
You don't have generalized drift detection yet. The audit is valuable precisely because it answers a narrower, answerable question — and that answer is what Mneme actually does today.