Project
Aftermath
Execution-backed verification layer for coding agents — durable receipts over narration.
active · ai / developer-tools · TypeScript
Research loop
- QUESTION
- RESEARCH PLAN
- SOURCES
- EVIDENCE
- CLAIMS
- SUPPORT / CONFLICT
- VERIFICATION
- ANSWER
Overview
Execution-backed verification layer for coding agents — durable receipts over narration.
Why it exists
Agents are persuasive. Reviewers need receipts: builds, tests, API drift, dependency changes — not self-grades.
Problem
Agent-authored code is often reviewed as prose. Confidence masquerades as evidence.
Approach
Independent examination of diffs and repo health signals, producing human- and machine-readable verification receipts.
Architecture
- Independent verification pass beside the agent
- Checks across build, test, lint, types, formatters, smoke, API drift
- Durable receipt artifacts for PR and automation consumption
Current status
Active open-source verification tooling for agent workflows.
Related
StackglassObservability and verification layer for AI-assisted development — evidence instead of guesses.CartographerTypeScript-native navigation system for software architecture — know where you are in the code.ArcframeLocal-first repository intelligence for Cursor MCP — graph, blast radius, evidence-backed context.