Power your agent
การติดตั้งจากผู้เผยแพร่ ไม่ใช่เอนด์พอยต์ของ PlugMCP
URL แบบรีโมตหรือคำสั่งบนเครื่องของคุณ? การติดตั้งสองแบบต่างกันอย่างไร
รันบน PlugMCP
A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.
เปิดStop wiring six MCP servers. Argyph is one local binary that gives your AI coding agent grep, a symbol graph, and semantic search over any repo — indexed in under a second, with no API key.
Install · Quick start · How it works · Tools · Architecture · Roadmap
⭐ If this looks useful, a star helps others find it.
Most developers stitch together a half-dozen MCP servers to give an AI agent context over a codebase — one for grep, one for embeddings, one for symbol search, one for repo packing. Each is a separate install, a separate process, and (often) a separate cloud account. Argyph replaces all of them with a single local binary. It indexes incrementally and is ready in under a second on previously-indexed repos — no daemon, no account, no API key.
The name is a portmanteau of Argus (the hundred-eyed watcher of Greek myth) and Glyph (a carved symbol with bound meaning) — a server that watches a codebase and gives the agent its symbols.
Argyph exposes three pillars of context behind one MCP endpoint:
[!NOTE] Everything is read-only. Argyph never edits, commits, or executes code.
Local-first by design. Every tier runs on the developer's machine — your proprietary code never touches a cloud vector DB or a remote embedding API, and a bundled embedding model means full functionality with zero keys.
Most existing context servers depend on a cloud vector database (Milvus, Pinecone) and a remote embedding API. That's a non-starter for proprietary code at most companies, and a tax on cold starts everywhere else.
Argyph runs entirely on the developer's machine: a single binary, an embedded vector store, an optional bundled embedding model — no daemon, no account, no key required to get full functionality.
Current release:
v1.0.4— published on npm, crates.io, the Homebrew tap, and as prebuilt GitHub-release binaries.
Claude Code (recommended):
claude mcp add argyph -- npx argyph@latest
npm / npx:
npx argyph
brew install Ezzy1630/argyph/argyph
curl -fsSL https://raw.githubusercontent.com/Ezzy1630/argyph/main/scripts/install.sh | bash
Intel Mac (
x86_64-apple-darwin) users: the ONNX Runtime backend doesn't ship a prebuilt for Intel macOS, so no prebuilt binary is produced for that target. Use the Cargo path below — it builds in a couple of minutes and works fine on Intel Macs.
cargo install argyph builds Argyph from source. Requires a Rust
toolchain ≥ 1.88 (rustup update). The bundled local embedder uses
ort (ONNX Runtime); on most
platforms ort ships a prebuilt dynamic library and there is nothing
to do. On Linux you may need libssl-dev and a working C toolchain;
on Windows the MSVC build tools are required. If the build fails on
ort-sys linkage, set ORT_STRATEGY=download before re-running.
cargo install argyph --locked
git clone https://github.com/Ezzy1630/argyph.git
cd argyph
cargo build --release
./target/release/argyph serve
Download argyph.dxt from the latest release and double-click.
# In any repo
cd ~/code/your-repo
argyph init
claude mcp add argyph -- npx argyph@latest
claude
In the chat:
What does this codebase do, and where is session expiration controlled?
Behind the scenes the agent calls ask, and Argyph returns bounded spans — not whole files — so the agent reads only what matters:
{
"spans": [
{
"file": "src/auth/session.rs",
"start_line": 41,
"end_line": 55,
"text": "fn is_expired(&self, now: Instant) -> bool { ... }",
"kind": "definition",
"symbol": "is_expired",
"language": "rust"
}
],
"strategy_used": "definition"
}
Argyph indexes Tier 0 in under a second on first run, Tier 1 (symbol graph) in seconds, and Tier 2 (embeddings) in the background. You can query immediately — tools return what's available now plus an index_coverage field so the agent knows.
Argyph builds the index in three tiers, each useful before the next completes:
The 80/20 insight: most agent queries are structural (where is parseConfig defined?, what calls validateUser?) and don't need embeddings. Argyph serves those from Tier 1 in milliseconds, even while Tier 2 is still building.
After the first index, the on-disk .argyph/ directory persists and only changed files are reprocessed on subsequent runs. A filesystem watcher keeps everything live.
Full schema reference: docs/tools-reference.md.
locate_smart is an opt-in tool that runs a bounded multi-step retrieval loop using an LLM provider. It's disabled by default; enable it in .argyph/config.toml:
[locate_smart]
enabled = true
provider = "openai" # or "anthropic" | "ollama"
model = "gpt-4o-mini"
# endpoint = "http://localhost:11434" # only for local providers
Or via env:
ARGYPH_LOCATE_SMART_ENABLED=1
ARGYPH_LOCATE_SMART_PROVIDER=anthropic
ARGYPH_LOCATE_SMART_MODEL=claude-haiku-4-5
When disabled, calls to locate_smart return LOCATE_SMART_DISABLED immediately and no provider keys are required.
Build with the smart feature:
cargo install argyph --features smart
Config is layered (highest priority first): env vars, .argyph/config.toml in the repo, built-in defaults. A config file is never required.
ARGYPH_LOG=info
ARGYPH_EMBED_PROVIDER=local # local | openai | voyage
OPENAI_API_KEY=... # standard provider env vars
ARGYPH_DISABLE_WATCHER=true # for sandboxed environments
Install agent lookup instructions in CLAUDE.md, AGENTS.md, or GEMINI.md:
argyph init
Argyph is the only one that's all five — symbol graph and semantic search, local-first, one install, incremental.
Full methodology and reproduction steps are in docs/benchmarks.md — reproduce locally with cargo bench --workspace. Numbers below are the median of three runs on the reference hardware tagged m4-air (Apple M4 MacBook Air, 16 GB RAM, macOS 26) unless stated otherwise.
Reproduce via:
cargo run --release -p argyph-benches --bin system_bench -- /path/to/repo
Tier 0 — the "useful immediately" gate — scales linearly and fast
(81K files in 2.1 s). search_text and pack_repo are available the
moment Tier 0 is ready.
Tier 1 (the symbol graph) is fast on normal-to-large repos. Two
optimizations landed for v1.0: the within-file reference resolver was
rewritten from an O(symbols²) substring scan to one-pass hash-indexed
lookups (4.2× — TypeScript compiler source: 34.6 s → 8.2 s), and the
per-file symbol/chunk SQL writes are now batched. On very large
monorepos (the full 81K-file TypeScript repo, ~2M LOC) Tier 1 still
does not finish within 30 min — the residual cost is raw tree-sitter
parse volume across 79K files, which needs streaming/parallel indexing
(tracked in ROADMAP.md). The server never blocks at any
scale: Tier-1 tools return INDEX_NOT_READY with a retry hint until
the graph is ready. Full data: docs/benchmarks.md.
Argyph is a Rust workspace of twelve focused crates with strict module ownership. The full architecture — including the Supervisor lifecycle, the three-tier indexing model, and per-crate responsibility boundaries — is documented in ARCHITECTURE.md.
Stable. v1.0.x is published on all channels. Active development continues against the milestones at the top of ROADMAP.md; the build plan is in docs/BUILD_PLAN.md.
Argyph is built with substantial AI assistance, but human-architected and human-reviewed. Contribution guide and the (strict) AI agent rules are in CONTRIBUTING.md. Commit conventions, including the project's attribution policy, are in docs/COMMIT_CONVENTIONS.md.
Built by Ezzy1630. See AUTHORS.md.
Dual-licensed under either of:
at your option.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase. Zero config, zero cloud, full context.
เครื่องมือ
ยังไม่มีรายการเครื่องมือ เซิร์ฟเวอร์แบบรีโมตจะได้รายการหลังการตรวจสอบสถานะครั้งแรก
FAQ
คัดลอกคอนฟิก stdio สำหรับ Cursor, Claude Desktop, Claude Code หรือ VS Code ได้ในหน้านี้ คอนฟิกนี้ใช้การติดตั้งของผู้เผยแพร่ ไม่ใช่พร็อกซีของ PlugMCP
ระเบียนนี้ไม่ได้ระบุตัวแปรสภาพแวดล้อมที่จำเป็น
Argyph ทำงานแบบโลคัลผ่าน stdio
ตรวจสอบล่าสุดเมื่อ 2026-10-28 Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase. Zero config, zero cloud, full context.