graqle 0.35.1

Give your AI tools architecture-aware reasoning. Build a knowledge graph from any codebase — dependency analysis, impact analysis, governed AI answers with confidence scores. Works with Claude Code, Cursor, VS Code Copilot. 14 LLM backends, fully offline capable.

Your AI reads files. GraQle reads architecture.

The context layer for AI coding agents. Scan any codebase into a knowledge graph. Every module becomes an agent. Ask questions — get architecture-aware answers in 5 seconds, not 2 minutes.

pip install graqle && graq scan repo . && graq run "what's the riskiest file to change?"

Website · Dashboard · PyPI · Changelog

---

50,000 tokens → 500 tokens. Same answer.

	Without GraQle	With GraQle
"What depends on auth?"	AI reads 60 files, guesses	Graph traversal → exact answer in 5s
Tokens per question	50,000	500
Cost per question	~$0.15	~$0.0003
Impact analysis	Manual grep + hope	graq impact auth.py → full blast radius
Memory across sessions	Lost when chat resets	Persistent knowledge graph
Confidence in answers	"I think..."	Confidence score + evidence chain

"We scanned 17,418 nodes across 3 projects in one session. Found 807 jargon blind spots, 218 ghost UI elements, and a CTA that was 20px tall (44px minimum). Cost: $0.30." — Quantamix Website Audit

---

How it works — 60 seconds

# 1. Install
pip install graqle

# 2. Scan your codebase into a knowledge graph
graq scan repo .
# → 2,847 nodes, 9,156 edges — your entire architecture mapped

# 3. Ask anything about your architecture
graq run "explain the payment flow end to end"
# → Graph-of-agents activates 8 relevant nodes, synthesizes answer
# → Confidence: 92% | Cost: $0.001 | Time: 5.2s

# 4. Connect to your AI IDE (zero config change)
graq init          # Claude Code, Cursor, VS Code, Windsurf — auto-detected

Your AI now has 27 architecture-aware MCP tools — including Phantom computer skills for live browser automation. No workflow change — it uses them automatically.

---

What makes Graqle different

🔬 Graph-of-Agents Reasoning Every module in your codebase becomes an autonomous agent. When you ask a question, only the relevant agents activate — they debate, exchange evidence, and synthesize one answer with a confidence score and full audit trail. This is not RAG. This is structured multi-agent reasoning over your dependency graph.	🧠 The Graph Learns graq learn "auth requires refresh token rotation" graq grow # Auto-runs on git commit Every interaction makes the graph smarter. Lessons persist across sessions. New developers and AI tools inherit your team's institutional knowledge automatically.
🛡️ Governed AI Decisions graq preflight "refactor the database layer" # → 4 modules depend on connection pool # → 2 have no tests # → DRACE score: 0.72 (proceed with caution) Every answer is auditable. DRACE governance scores across 5 dimensions. Full evidence chains. Patent-protected.	⚡ 14 LLM Backends model: backend: ollama # Free, offline, air-gapped # Also: anthropic, openai, groq, deepseek, # gemini, bedrock, together, mistral, # fireworks, cohere, openrouter, vllm, custom Use your own API keys. Run fully offline with Ollama. Smart routing assigns different models to different tasks.

---

Real stories from production

📊 "807 jargon blind spots in 90 seconds" — Website audit with SCORCH

A professional website with WCAG AAA compliance still had 807 unexplained acronyms (TAMR+, TRACE, SHACL, HashGNN) that compliance officers would bounce on. GraQle's SCORCH engine found them all in one scan. Lighthouse missed every one.

Before: "Explore our TAMR+ SHACL-compliant governance pipeline" After: "Explore our regulatory compliance pipeline" (with inline tooltips for technical terms)

🏗️ "17,418 nodes, 8 audits, $0.30" — Multi-project knowledge graph

Three repos merged into one knowledge graph. 8 parallel audits ran across the entire surface. Found a CTA button that was only 20px tall (44px minimum for mobile touch targets). Fixed before a single prospect saw it.

Scale: 17,418 nodes | 70,545 edges | 8 audits | Total cost: $0.30

🎯 "From score 12 to production-ready in one night" — Canvas workflow audit

GraQle's SCORCH engine audited a complete Canvas workflow builder. Initial score: 12/100. After one session of GraQle-guided fixes: production-ready. Zero manual testing — the graph knew which components to check and in what order.

🔄 "6.4 → 8.5 across 5 releases" — SDK dogfooding journey

GraQle scores itself on every release. From v0.12.3 (6.4/10) to v0.29.9 (8.5/10) — every improvement was guided by the knowledge graph's own intelligence layer. 2,000+ tests. 396 compiled modules. Graph-powered development, by the graph.

---

NEW: Phantom — Computer Skills for AI Agents

Your AI can now open a real browser, click buttons, fill forms, and audit any website.

Phantom is GraQle's browser automation plugin. It gives Claude Code, Cursor, and any MCP-compatible AI the ability to interact with live web applications — not just read code, but use the product.

pip install graqle[phantom] && python -m playwright install chromium

Capability	What it does
Browse	Open any URL, get screenshot + full DOM summary
Click / Type	Interact with elements by text, CSS selector, or coordinates
Audit	Run 10 dimensions (a11y, security, mobile, SEO, brand, performance, ...)
Flow	Execute multi-step user journeys with assertions
Discover	Auto-crawl all pages from a starting URL
Vision	Claude Vision analyzes screenshots for UX friction
Learn	Auto-record critical findings into the knowledge graph

# Audit any website — 10 dimensions, one command
graq phantom audit https://example.com

# Discover all pages
graq phantom discover https://example.com

# Run a complete user journey from a JSON flow file
graq phantom flow journey.json

Product-agnostic. Works on any website. No configuration needed. Results feed back into the GraQle knowledge graph so your AI learns from every audit.

---

IDE integration — one command

graq init                    # Claude Code — auto-wires MCP tools
graq init --ide cursor       # Cursor — MCP + .cursorrules
graq init --ide vscode       # VS Code + Copilot
graq init --ide windsurf     # Windsurf — MCP + .windsurfrules

27 MCP Tools

Tool	What it does	Free
graq_context	Focused 500-token context for any module	✅
graq_reason	Multi-agent graph reasoning	✅
graq_impact	Blast radius — what breaks if you change this	✅
graq_preflight	Pre-change safety check with risk scoring	✅
graq_lessons	Relevant lessons from past mistakes	✅
graq_learn	Teach the graph new knowledge	✅
graq_gate	Governance gate (pass/fail)	✅
graq_drace	DRACE quality score (5 dimensions)	✅
graq_scorch_audit	Full UX friction audit (Claude Vision)	Pro
graq_scorch_behavioral	12 behavioral UX tests (free, no AI)	✅
graq_phantom_browse	Open browser, navigate, screenshot + DOM	Pro
graq_phantom_click	Click elements by text, selector, or coordinates	Pro
graq_phantom_type	Type into forms, search boxes, inputs	Pro
graq_phantom_audit	Run 10 audit dimensions on any live page	Pro
graq_phantom_flow	Execute multi-step user journeys	Pro
graq_phantom_discover	Auto-discover all pages from a starting URL	Pro
+11 more	inspect, reload, route, screenshot, session...	✅

---

Architecture

Your Code                    Knowledge Graph               AI Reasoning
┌─────────────┐             ┌──────────────────┐          ┌─────────────────┐
│ Python      │  graq scan  │  Nodes (modules) │  query   │ Graph-of-Agents │
│ TypeScript  │ ──────────> │  Edges (depends) │ ───────> │ Multi-round     │
│ Config      │             │  Skills (201)    │          │ Confidence-scored│
│ Docs        │             │  Lessons         │          │ Audit-trailed   │
└─────────────┘             └──────────────────┘          └─────────────────┘
                                    │
                              graq learn / graq grow
                                    │
                            Graph evolves with
                            every interaction

Languages: Python, JavaScript/TypeScript, React/JSX, Go, Rust, Java Frameworks: FastAPI, Django, Flask, Next.js, React, Express, NestJS Documents: PDF, DOCX, PPTX, XLSX, Markdown

---

Full CLI reference

Scan & Build

Command	Description
graq init	Scan repo, build graph, auto-wire IDE
graq scan repo .	Scan codebase into knowledge graph
graq scan docs ./docs	Ingest documents into the graph
graq compile	Risk scores, insights, CLAUDE.md auto-injection
graq verify	Run governance gate checks

Query & Reason

Command	Description
graq run "question"	Natural language query (auto-routed)
graq reason "question"	Multi-agent graph reasoning
graq context module-name	Focused 500-token context
graq impact module-name	Downstream impact analysis
graq preflight "change"	Pre-change safety check
graq lessons topic	Surface relevant lessons
graq predict "query"	Confidence-gated prediction + CI gate

Teach & Learn

Command	Description
graq learn "fact"	Teach the graph knowledge
graq learn node "name"	Add a node
graq learn edge "A" "B"	Add a relationship
graq learned	List what the graph knows
graq grow	Incremental update (git hook)

Cloud & Infrastructure

Command	Description
graq login --api-key grq_...	Authenticate
graq cloud push	Upload graph to cloud
graq cloud pull	Download graph
graq studio	Visual dashboard
graq serve	REST API server
graq mcp serve	MCP server for IDEs
graq doctor	Health check
graq self-update	Upgrade GraQle

SCORCH — UX Friction Auditing (13 tests)

Command	Description
graq scorch run	Full 5-phase audit
graq scorch behavioral	12 behavioral tests (free)
graq scorch a11y	Accessibility (WCAG 2.1)
graq scorch perf	Core Web Vitals
graq scorch seo	SEO + Open Graph
graq scorch mobile	Touch targets + viewport
graq scorch security	CSP, XSS, exposed keys
graq scorch conversion	CTA + trust signals
graq scorch brand	Visual consistency

Phantom — Browser Automation + Computer Skills (8 tools)

Command	Description
graq phantom browse URL	Open browser, screenshot + DOM summary
graq phantom audit URL	Run 10 audit dimensions on any page
graq phantom discover URL	Auto-discover all navigable pages
graq phantom flow file.json	Execute multi-step user journey

MCP tools (used automatically by your AI):

Tool	Description
graq_phantom_browse	Navigate to URL, return screenshot + DOM
graq_phantom_click	Click by text, CSS selector, or coordinates
graq_phantom_type	Type into forms and inputs
graq_phantom_screenshot	Capture + optional Claude Vision analysis
graq_phantom_audit	10-dimension audit on live page
graq_phantom_flow	Multi-step user journey execution
graq_phantom_discover	Auto-crawl all pages from starting URL
graq_phantom_session	Session + auth profile management

Requires: pip install graqle[phantom] && python -m playwright install chromium

---

14 LLM Backends

Backend	Best For	Cost
Ollama	Offline, air-gapped, privacy	$0
Groq	Speed — sub-second responses	~$0.0005/q
DeepSeek	Budget-conscious	~$0.0001/q
Anthropic	Complex reasoning	~$0.001/q
OpenAI	Broad compatibility	~$0.001/q
Google Gemini	Long context	~$0.0001/q
AWS Bedrock	Enterprise IAM	AWS pricing
Together / Mistral / Fireworks / Cohere / OpenRouter / vLLM / Custom	Various	Various

# graqle.yaml — smart routing
routing:
  default_provider: groq          # Fast for lookups
  rules:
    - task: reason
      provider: anthropic         # Claude for deep reasoning

---

Pricing

	Free ($0)	Pro ($19/mo)	Team ($29/dev/mo)
CLI + SDK + MCP	Unlimited	Unlimited	Unlimited
All 14 backends	✅	✅	✅
Graph nodes	500	25,000	Unlimited
Cloud projects	1	3	Unlimited
SCORCH Vision	—	✅	✅
Phantom Computer Skills	—	✅	✅
Cross-project	—	✅	✅
Team graphs	—	—	✅

Start free →

---

Security & Privacy

• Local by default. All processing runs on your machine.
• No telemetry. GraQle does not phone home.
• Your API keys. LLM calls go directly to your provider.
• Cloud is opt-in. Uploads graph structure only — never source code.

Supply-Chain Integrity (v0.35.0+)

Every Graqle release is hardened against supply-chain attacks:

Protection	What it does
PyPI Trusted Publishing	No long-lived API tokens — releases are tied to GitHub Actions OIDC
Sigstore signatures	Every wheel is signed; bundle attached to each GitHub Release
CycloneDX SBOM	Full bill of materials for every release
pip-audit in CI	CVE scan on every PR — blocks on CRITICAL/HIGH
.pth file guard	Publish is blocked if wheel contains .pth files (LiteLLM-class attack prevention)
Reproducible builds	SOURCE_DATE_EPOCH pinned — rebuild and compare checksums

Verify any release in one command:

pip install "graqle[security]"
graq trustctl verify                    # verify installed version
graq trustctl verify --version 0.35.0  # verify a specific release

Use in your CI pipeline (see tools/verify-graqle-example.yml for the full template):

- name: Verify Graqle integrity
  run: |
    pip install "graqle[security]==0.35.0"
    graq trustctl verify --version 0.35.0

See SECURITY.md for the full disclosure policy and supply-chain documentation.

---

FAQ

Why not just use Cursor / Claude Code / Copilot directly?

GraQle doesn't replace your AI tool — it makes it 100x better. Your AI reads files one at a time and guesses at relationships. With GraQle, it queries a knowledge graph that maps your entire architecture. Same AI, 100x fewer tokens, answers grounded in actual dependency structure. Plugs in via MCP with zero workflow change.

How is this different from Sourcegraph or static analysis?

Static analysis tells you what code exists. GraQle tells you how it connects, what breaks when it changes, and what your team has learned about it. Every answer comes with a confidence score and evidence chain. It's a reasoning layer, not a search engine.

Does my code leave my machine?

Never. All processing is local. Cloud sync uploads graph structure only — never source code.

Can I use my own LLM?

Yes. 14 backends including Ollama for fully offline operation. Any OpenAI-compatible endpoint works.

How long does scanning take?

Under 30 seconds for most codebases. 10K+ file monorepos take 1-2 minutes.

---

Patent & License

European Patent Applications EP26162901.8 and EP26166054.2 — Quantamix Solutions B.V. Phantom browser automation plugin: Copyright 2026 Quantamix Solutions B.V. All rights reserved. Free to use under the license terms. See SECURITY.md.

@article{kumar2026graqle,
  title   = {GraQle: Governed Intelligence through Graph-of-Agents Reasoning},
  author  = {Kumar, Harish},
  year    = {2026},
  institution = {Quantamix Solutions B.V.},
  url     = {https://github.com/quantamixsol/graqle}
}

---

Your AI is only as good as the context you give it. Give it your architecture.

pip install graqle && graq init

⭐ If GraQle saved you time, star this repo — it helps other developers find it.

Built by Quantamix Solutions B.V. — Uithoorn, The Netherlands 🇳🇱