Item detail

archcore-ai/cli

Archcore is a CLI and local stdio MCP server that gives your AI coding agent persistent, Git-versioned memory of your architecture decisions, rules, conventions, plans, and incidents. 47 stars, Apache-2.0, Go, works across Claude Code, Cursor, Copilot, Gemini CLI, Codex, OpenCode, Roo Code, and Cline. Stores decisions as structured ADRs, rules, plans, and incident docs in `.archcore/` so any futur

Score7.5
Popularity60.0
Risknone
TierSilver
Score breakdown
Usefulness7.0
Novelty9.0
Momentum5.0
Maturity6.7
Open-source/build8.4
Evidence7.2
Workflow potential8.3
Setup ease8.8

Popularity is tracked separately. Support, ads, sponsorships, and tips never affect these signals.

Why it matters

Useful for engineering teams who want their AI coding agent to follow their repo's actual architecture, conventions, and past decisions — instead of guessing from code examples — across every tool (Claude Code, Cursor, Copilot, Codex, etc.).

Who should use it

engineering teams using AI coding agents that repeatedly violate the same repo conventionsplatform teams that need ADRs, rules, and incident docs to be readable by every AI tool (not just one)multi-agent-stack teams (Claude Code + Cursor + Copilot) who want one context layer across all of themengineering managers who want AI-assisted PRs to follow the same conventions human reviewers enforceOSS maintainers who want AI agent contributions to follow their contribution guidelines automatically

Who should skip it

Skip if the source link, docs, or setup requirements do not match your workflow.

Risk explanation

47 stars as of 2026-06 — fresh bet, plan for API / schema changes as the project matures; Distinct from agnix (cycle-23, linter) — Archcore is the enrichment layer, not the validation layer; pair them for defense in depth.

Evidence links

Closest alternatives / related signals

ai-coding-agentcontext-layeradrmcp-serverclaude-codecursorcopilotcodex