By anortham
Local SQLite-backed code intelligence, content search, and workspace freshness for coding agents. Requires Node.js on PATH (the plugin launcher is a Node script).
Use when resuming Miller-backed work from a handoff packet or validating a packet before continuing in another harness, model, or session.
Use when preparing to move active Miller-backed work to another harness, model, or session and the receiving agent needs a self-contained resume packet.
Use when tracing Miller bridge paths in supported providers: dotnet-web URL literals and fetch/axios client requests to ASP.NET signals, client requests to Next.js route handlers or Nuxt server routes, Next.js/Nuxt route references to file routes, Vue/React references to route definitions, or backend-http client requests to Express/FastAPI/Flask/Django/Spring/Go/gin/echo/Rails/NestJS/Laravel/Phoenix/axum/actix/Symfony/Ktor route facts.
Use when querying another registered workspace with Miller, comparing repos, opening a workspace, or routing search/inspect/context/impact/trace through workspace_id.
Use before changing existing indexed files with Miller edit, especially symbol rewrites, text replacements, renames, or refactors.
Admin access level
Server config contains admin-level keywords
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Executables (bin/) — files in this plugin's bin/directory are added to the Bash tool's PATH while the plugin is enabled.
Miller is a local code-intelligence server for coding agents. It keeps a current SQLite-backed view of a workspace, then answers structural questions through MCP and a matching CLI: find symbols, inspect files, build focused context, trace relationships, assess change impact, and check workspace freshness without asking an agent to grep and reread the repo by hand.
Miller is the free local core in the Miller/Eros product split. It stays deterministic, lexical/structural, daemon-light, and local-first, with optional local semantic retrieval that is off-switchable and leaves lexical-only results byte-identical. Eros sits above it for fleet-level semantics: cross-workspace ranking, higher-level guidance, confidence/evidence views, and commercial orchestration.
The practical difference from a one-time graph dump is that Miller is built for active agent work:
MILLER_SEARCH_SIDECAR=0 when debugging the in-memory fallback;Current release: v1.13.0. Miller ships as agent plugins, self-contained per-platform release archives, and a source-checkout workflow. Plugin and release-archive installs include the pinned
julie-extractbinary; users do not install it separately.Website: anortham.github.io/miller · Release: v1.13.0
Most Claude Code and Codex users should start with the agent plugin. The plugin launcher downloads the matching
Miller release archive, verifies its .sha256 sidecar, caches it under ~/.miller/plugin-cache/, and starts
miller serve as an MCP server. That archive includes Miller's pinned .tools/julie-extract binary, so plugin
users do not install julie-extract or the .NET SDK separately.
Plugin installs require Node.js on
PATH— the launcher is a Node script. If Node.js is missing (common with Claude Code's native installer, which does not need Node itself), the plugin fails to connect with the opaque MCP error-32000and writes no Miller log. Install Node.js LTS and restart your agent. The manual install paths below run themillerbinary directly and do not need Node.js.
Claude Code:
/plugin marketplace add anortham/miller
/plugin install miller@miller
Codex:
codex plugin marketplace add anortham/miller
codex
# then open /plugins and install Miller from the miller marketplace
Cursor: install Miller from the Cursor plugin marketplace, or add a user-global ~/.cursor/mcp.json
entry with an absolute path to miller serve. Miller binds its workspace from MCP client roots on the first
tool call, so one global install works per editor window without ${workspaceFolder} placeholders.
The Claude Code plugin also injects a ~2.4KB Miller routing block at session start through a SessionStart hook,
so the tool-routing guidance stays in context even though clients truncate MCP server instructions. The Codex
plugin ships the same hook: Codex runs it once you review and trust the plugin's hooks, and current Codex builds
load hooks from ~/.codex/hooks.json rather than from plugin roots
(openai/codex#16430). Set MILLER_SESSION_HOOKS=0 to opt out.
After extracting a release archive (see Manual Binary Install), add to
~/.cursor/mcp.json:
{
"mcpServers": {
"miller": {
"type": "stdio",
"command": "/absolute/path/to/miller-1.13.0-aarch64-apple-darwin/miller",
"args": ["serve"]
}
}
}
On Windows, use the full path to miller.exe as command. Optional override for clients without MCP roots:
set "env": { "MILLER_WORKSPACE_ROOT": "/absolute/path/to/project" } on the server entry.
For Miller development in this checkout, run scripts/install-cursor-local-dev.sh (or .ps1) to write
~/.cursor/mcp.json with your Release build path and retire legacy plugin-cache installs.
Reload Cursor after adding the entry, then ask your agent to search, inspect, build context, trace, or check impact
with Miller. Miller writes its local index under that workspace's .miller/ directory.
Use Miller's structural tools before broad file reads:
npx claudepluginhub anortham/miller --plugin millerMiller-powered development workflow with subagent-driven execution, inline review, and Miller-first codebase orientation.
Cross-client MCP memory with checkpoints, recall, briefs, and standups for AI-assisted development
Code intelligence server: search, navigation, and refactoring across 34 languages
Agent memory persistence - remember and recall across sessions
Composite workflows that orchestrate Julie (code intelligence), Sherpa (workflow guidance), and Goldfish (persistent memory) for powerful development patterns. Includes TDD Powerhouse, Bug Detective, Smart Session Start, and more.
Intelligent draw.io diagramming plugin with AI-powered diagram generation, multi-platform embedding (GitHub, Confluence, Azure DevOps, Notion, Teams, Harness), conditional formatting, live data binding, and MCP server integration for programmatic diagram creation and management.
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Lazy senior dev mode. Forces the simplest, shortest solution that actually works: YAGNI, stdlib first, no unrequested abstractions.
Access thousands of AI prompts and skills directly in your AI coding assistant. Search prompts, discover skills, save your own, and improve prompts with AI.
Consult multiple AI coding agents (Gemini, OpenAI, Grok, Perplexity, plus codex, antigravity, and grok CLIs when installed) to get diverse perspectives on coding problems