![\"\"](\"https://raw.githubusercontent.com/Karzone/TestAtlas/main/assets/logo-mark.svg\"){kind=logo}
\n TestAtlas\n\n TestAtlas\n\n A queryable, semantic map of your .NET test-automation solution — in one SQLite file, served to your AI agent over MCP.\n
\n\n\n Zero config · No AI · No network · Deterministic\n
\n\n\n ![\"TestAtlas.Mcp](\"https://img.shields.io/nuget/v/TestAtlas.Mcp?logo=nuget&label=TestAtlas.Mcp&color=004880\")
![\"TestAtlas.Cli](\"https://img.shields.io/nuget/v/TestAtlas.Cli?logo=nuget&label=TestAtlas.Cli&color=004880\")
![\"Model](\"https://img.shields.io/badge/MCP_Registry-listed-7C3AED\"){kind=icon}
![\"Glama](\"https://img.shields.io/badge/Glama-listed-1B8F5A\"){kind=icon}
![\".NET](\"https://img.shields.io/badge/.NET-8.0-512BD4?logo=dotnet&logoColor=white\"){kind=icon}
![\"MIT](\"https://img.shields.io/badge/license-MIT-blue\"){kind=icon}
\n \n
\n 11 MCP tools
\n resolve_step · step_catalog · impact · search_steps · search_scenarios · get_scenario · get_step_definition · list_tags · list_endpoints · project_dependencies · stats\n
\n The problem ·\n See it in action ·\n Quick start ·\n MCP ·\n Commands ·\n Fresh maps ·\n Roadmap\n
\n\n---\n\n## The problem\n\nLarge test-automation solutions are hard to navigate — for humans *and* for AI agents. Asked to\nautomate a new story, an agent can't see which steps already exist, where similar code lives, or\nwhat conventions the solution follows — so it duplicates steps and misplaces code. TestAtlas\nindexes the solution once into a single SQLite map and answers those questions precisely:\ndeterministically, offline, without a model or a network call.\n\n---\n\n## See it in action\n\nIndex the bundled 8-project sample once — `testatlas index samples/SampleShop/SampleShop.sln` — then\nask it questions from the terminal. Every number below is real output from that run:\n\n```console\n$ testatlas stats sampleshop.db\nTestAtlas map: sampleshop.db (schema v5)\n\ntotals: 8 project(s), 15 class(es), 36 method(s), 14 step definition(s)\nclass kinds:\n api_client 7\n page_object 4\n step_class 4\n\ngherkin: 4 feature(s), 5 scenario(s), 16 step(s)\nbound steps: 16 · unbound: 0 · ambiguous: 0\nendpoints: 3 (3 call site(s))\n```\n\n…or let your agent ask them over MCP. Here an agent checks whether a step it is about to write\nalready exists — `resolve_step` answers with the definitions that would bind it, or the closest\nnear-misses to reuse instead:\n\n```jsonc\n// agent → testatlas: resolve_step { \"text\": \"I add the product to my cart\" }\n{\n \"status\": \"none\", // nothing binds this exact text — don't invent it from scratch:\n \"suggestions\": [ // these existing steps are the closest, ranked by shared terms\n { \"expression\": \"product (.*) is added to the cart with quantity (.*)\",\n \"keyword\": \"When\", \"location\": \"SampleShop.Tests.Api/Steps/CatalogApiSteps.cs:34\" },\n { \"expression\": \"the cart is not empty\",\n \"keyword\": \"Then\", \"location\": \"SampleShop.Tests.Api/Steps/CatalogApiSteps.cs:43\" }\n // …8 more\n ]\n}\n```\n\nLive sample outputs, committed from that same solution:\n**[HTML report](https://htmlpreview.github.io/?https://github.com/Karzone/TestAtlas/blob/main/docs/sample-report.html)** (features, scenarios, bindings, class kinds, endpoints) ·\n**[dependency map](https://htmlpreview.github.io/?https://github.com/Karzone/TestAtlas/blob/main/docs/sample-map.html)** (the eight projects and their edges).\n(GitHub serves `.html` as source, so the links route through htmlpreview.github.io — or download from [`docs/`](docs/) and open locally.)\n\n## What you get\n\nTestAtlas statically analyses the solution and emits `codemap.db` — projects and their dependency\nedges, Gherkin features/scenarios/steps, step definitions and their bindings (bound / unbound /\nambiguous), page objects, API clients, helpers, test classes, and the call/usage edges connecting\nthem — then turns that map into answers:\n\n| | Capability | What you get |\n|:--:|---|---|\n| 🧩 | **Reuse-first authoring** | `resolve_step` — would this phrase bind an existing definition? `step_catalog` — the reusable step vocabulary with placeholders + allowed values |\n| 💥 | **Impact** | Blast radius — the scenarios affected by changing a class, method, step, or endpoint |\n| 🔍 | **Search** | FTS5 over step definitions + scenarios — *\"does a step for this already exist?\"* |\n| 🔌 | **MCP** | All of it served to an AI agent over stdio — precise answers in a few hundred tokens, no context stuffing |\n| 📊 | **Report & map** | Self-contained HTML drill-down of the whole map + project-dependency graph |\n| 📈 | **Stats** | Entity counts, class-kind breakdown, binding coverage, diagnostics |\n\nAll of it **offline, deterministic, and reproducible** — same input, same map, every time.\n\n---\n\n## 🚀 Quick start\n\n> **Requires** the [.NET SDK 8.0+](https://dotnet.microsoft.com/download). On a corporate machine\n> where `dotnet tool install` fails with **401**, see\n> [docs/troubleshooting.md](docs/troubleshooting.md).\n\n**1 — Install** the CLI (and the MCP server, if you'll connect an agent):\n\n```bash\ndotnet tool install --global TestAtlas.Cli\ndotnet tool install --global TestAtlas.Mcp\n```\n\n**2 — Index** your solution. This produces the map (`./codemap.db`) that every query, report, and\nMCP answer reads — nothing works without it:\n\n```bash\ntestatlas index path/to/YourSolution.sln\n```\n\nNo need to build or restore the solution first — indexing is a syntax-only pass, so an unrestored\ncheckout maps fine. Point `index` at a folder (or nothing) and it auto-discovers a single\n`.sln`/`.csproj` there.\n\n**3 — Query it:**\n\n```bash\ntestatlas stats\ntestatlas search \"login\"\ntestatlas report # writes codemap.html\ntestatlas map # writes codemap-map.html\n```\n\n…and to serve it to your AI agent, continue to [MCP setup](#-use-it-from-an-ai-agent-mcp).\n\n