cortexcode 0.9.0__tar.gz → 0.10.0__tar.gz

{cortexcode-0.9.0 → cortexcode-0.10.0}/PKG-INFO

@@ -1,7 +1,7 @@
 Metadata-Version: 2.4
 Name: cortexcode
-Version: 0.9.0
-Summary: Lightweight code indexing for AI assistants — save 90%+ tokens with structured context
+Version: 0.10.0
+Summary: Lightweight code indexing for AI assistants — build a searchable codegraph and grounded context
 Author-email: Naveen <naveen_joshi07@outlook.com>
 License: MIT
 Project-URL: Homepage, https://github.com/naveen-joshi/cortexcode
@@ -55,44 +55,36 @@ Dynamic: license-file
   <h1 align="center">CortexCode</h1>
   <p align="center">
     <strong>Lightweight code indexing for AI assistants</strong><br>
-    Save 90%+ tokens by giving AI agents structured context instead of raw source files.
+    Build a searchable codegraph, generate documentation, and give AI assistants grounded context.
   </p>
 </p>
 
-<p align="center">
-  <a href="https://pypi.org/project/cortexcode/"><img src="https://img.shields.io/pypi/v/cortexcode?style=flat-square&color=blue" alt="PyPI"></a>
-  <a href="https://pypi.org/project/cortexcode/"><img src="https://img.shields.io/pypi/pyversions/cortexcode?style=flat-square" alt="Python"></a>
-  <a href="https://github.com/naveen-joshi/cortexcode/blob/main/LICENSE"><img src="https://img.shields.io/badge/license-MIT-green?style=flat-square" alt="License"></a>
-  <a href="https://github.com/naveen-joshi/cortexcode"><img src="https://img.shields.io/github/stars/naveen-joshi/cortexcode?style=flat-square" alt="Stars"></a>
-  <a href="https://marketplace.visualstudio.com/items?itemName=cortexcode.cortexcode-vscode"><img src="https://img.shields.io/visual-studio-marketplace/v/cortexcode.cortexcode-vscode?style=flat-square" alt="VS Code"></a>
-</p>
-
 ---
 
 ## The Problem
 
 AI coding assistants (Copilot, Cursor, Windsurf, etc.) need to understand your codebase. The current approach: **dump entire source files into the context window**. This is:
 
-- **Expensive** — A 150-file project can cost 200K+ tokens per query
-- **Slow** — More tokens = slower responses
-- **Wasteful** — Most of those tokens are irrelevant to the question
+- **Noisy** — Important signals are buried inside large files
+- **Fragile** — Assistants miss architecture, relationships, and runtime surface area
+- **Hard to reuse** — Raw file dumps do not become a durable project map
 
 ## The Solution
 
-CortexCode indexes your codebase using **AST parsing** (tree-sitter) and provides a structured, searchable index. Instead of feeding 200K tokens of raw code, you feed **~500 tokens of relevant context**.
+CortexCode indexes your codebase using **AST parsing** (tree-sitter) and builds a structured, searchable codegraph. You can explore symbols, trace flows, generate docs and diagrams, produce CodeWiki pages, and connect AI tools through MCP.
 
 ```
 ┌─────────────────────────────────────────────────┐
 │  Without CortexCode          With CortexCode    │
 │                                                  │
-│  200,000 tokens    →         500 tokens          │
-│  $0.006/query      →         $0.00002/query      │
-│  All files dumped  →         Only relevant syms  │
-│  No structure      →         Call graph + types   │
+│  Raw files only    →         Searchable codegraph │
+│  Manual digging    →         Linked symbols       │
+│  Ad-hoc prompts    →         Reusable context     │
+│  No project map    →         Docs + diagrams      │
 └─────────────────────────────────────────────────┘
 ```
 
-Run `cortexcode stats` on your project to see your actual savings.
+Run `cortexcode index` on your project to generate the codegraph and start exploring.
 
 ## Quick Start
 
@@ -108,9 +100,6 @@ cd cortexcode && pip install -e .
 cd your-project
 cortexcode index
 
-# See token savings
-cortexcode stats
-
 # Get context for AI
 cortexcode context "handleAuth"
 
@@ -159,24 +148,14 @@ Parses source code into structured symbols using tree-sitter grammars.
 - **Entities** — Database models and ORM definitions
 - **Framework Detection** — React components, Angular services, etc.
 
-### Token Savings
+### Project Exploration & Context
 
-CortexCode dramatically reduces the tokens needed to give AI assistants useful context:
+Use CortexCode to inspect structure and answer practical questions about a codebase:
 
-```bash
-$ cortexcode stats
-
-╭──────── Token Savings Analysis ────────╮
-│ Source files         154 files          │
-│ Raw project tokens   203,847            │
-│ Full index tokens    45,291             │
-│ Context query tokens 487                │
-│                                         │
-│ Tokens saved         203,360            │
-│ Savings              99.8%              │
-│ Compression ratio    418.6x             │
-╰─────────────────────────────────────────╯
-```
+- **Context lookup** — Pull relevant symbols and files for a query
+- **Call graph tracing** — Follow how behavior moves across modules
+- **Architecture visibility** — Inspect routes, entities, dependencies, and layers
+- **Reusable index** — Keep a structured project map that tools can build on
 
 ### Interactive HTML Documentation
 
@@ -211,7 +190,6 @@ cortexcode wiki --no-modules       # Skip per-module pages (faster)
 - **Mermaid diagrams** — Auto-generated flow diagrams
 - **Concept mapping** — Maps technical concepts to symbols and files
 - **Concept search** — Ask "how does authentication work?" and get grounded answers
-- **Token tracking** — See exactly how many tokens each page used
 
 **Output:** `.cortexcode/wiki/index.html` — Open directly or serve locally.
 
@@ -262,12 +240,12 @@ CortexCode supports both a short form (`cc`) and full form (`cortexcode`). Comma
 | Command | Description |
 |---------|-------------|
 | `cc analyze context [query]` | Get relevant context for AI |
-| `cc analyze context [query] --tokens` | Show token savings for query |
+| `cc analyze context [query] --tokens` | Show context size estimates for a query |
 | `cc analyze search [query]` | Grep-like symbol search with type/file filters |
 | `cc analyze find [query]` | Semantic search by meaning ("auth handler") |
 | `cc analyze diff` | Show changed symbols since last commit |
 | `cc analyze diff --ref HEAD~3` | Compare against any git ref |
-| `cc analyze stats` | Show project stats and token savings |
+| `cc analyze stats` | Show project index and analysis stats |
 | `cc analyze scan` | Scan dependencies for security warnings |
 | `cc analyze trace <symbol>` | Trace code flow through call graph |
 | `cc analyze flow <concept>` | Analyze code flow grouped by file |
@@ -372,7 +350,7 @@ Add to `~/.windsurf/config.json`:
 
 ```bash
 # Paste this output into your AI chat
-cortexcode context "useAuth" --tokens
+cortexcode context "useAuth"
 ```
 
 ### 2. JSON Index (programmatic)
@@ -569,7 +547,6 @@ CortexCode respects `.gitignore` files (including nested ones) and has built-in
 ## Roadmap
 
 - [x] MCP server for direct AI agent integration
-- [x] Tiktoken-based accurate token counting
 - [x] Semantic search over symbols (TF-IDF + synonym expansion)
 - [x] Cross-file type inference
 - [x] Git diff-aware context (show only changed symbols)

{cortexcode-0.9.0 → cortexcode-0.10.0}/README.md

@@ -2,44 +2,36 @@
   <h1 align="center">CortexCode</h1>
   <p align="center">
     <strong>Lightweight code indexing for AI assistants</strong><br>
-    Save 90%+ tokens by giving AI agents structured context instead of raw source files.
+    Build a searchable codegraph, generate documentation, and give AI assistants grounded context.
   </p>
 </p>
 
-<p align="center">
-  <a href="https://pypi.org/project/cortexcode/"><img src="https://img.shields.io/pypi/v/cortexcode?style=flat-square&color=blue" alt="PyPI"></a>
-  <a href="https://pypi.org/project/cortexcode/"><img src="https://img.shields.io/pypi/pyversions/cortexcode?style=flat-square" alt="Python"></a>
-  <a href="https://github.com/naveen-joshi/cortexcode/blob/main/LICENSE"><img src="https://img.shields.io/badge/license-MIT-green?style=flat-square" alt="License"></a>
-  <a href="https://github.com/naveen-joshi/cortexcode"><img src="https://img.shields.io/github/stars/naveen-joshi/cortexcode?style=flat-square" alt="Stars"></a>
-  <a href="https://marketplace.visualstudio.com/items?itemName=cortexcode.cortexcode-vscode"><img src="https://img.shields.io/visual-studio-marketplace/v/cortexcode.cortexcode-vscode?style=flat-square" alt="VS Code"></a>
-</p>
-
 ---
 
 ## The Problem
 
 AI coding assistants (Copilot, Cursor, Windsurf, etc.) need to understand your codebase. The current approach: **dump entire source files into the context window**. This is:
 
-- **Expensive** — A 150-file project can cost 200K+ tokens per query
-- **Slow** — More tokens = slower responses
-- **Wasteful** — Most of those tokens are irrelevant to the question
+- **Noisy** — Important signals are buried inside large files
+- **Fragile** — Assistants miss architecture, relationships, and runtime surface area
+- **Hard to reuse** — Raw file dumps do not become a durable project map
 
 ## The Solution
 
-CortexCode indexes your codebase using **AST parsing** (tree-sitter) and provides a structured, searchable index. Instead of feeding 200K tokens of raw code, you feed **~500 tokens of relevant context**.
+CortexCode indexes your codebase using **AST parsing** (tree-sitter) and builds a structured, searchable codegraph. You can explore symbols, trace flows, generate docs and diagrams, produce CodeWiki pages, and connect AI tools through MCP.
 
 ```
 ┌─────────────────────────────────────────────────┐
 │  Without CortexCode          With CortexCode    │
 │                                                  │
-│  200,000 tokens    →         500 tokens          │
-│  $0.006/query      →         $0.00002/query      │
-│  All files dumped  →         Only relevant syms  │
-│  No structure      →         Call graph + types   │
+│  Raw files only    →         Searchable codegraph │
+│  Manual digging    →         Linked symbols       │
+│  Ad-hoc prompts    →         Reusable context     │
+│  No project map    →         Docs + diagrams      │
 └─────────────────────────────────────────────────┘
 ```
 
-Run `cortexcode stats` on your project to see your actual savings.
+Run `cortexcode index` on your project to generate the codegraph and start exploring.
 
 ## Quick Start
 
@@ -55,9 +47,6 @@ cd cortexcode && pip install -e .
 cd your-project
 cortexcode index
 
-# See token savings
-cortexcode stats
-
 # Get context for AI
 cortexcode context "handleAuth"
 
@@ -106,24 +95,14 @@ Parses source code into structured symbols using tree-sitter grammars.
 - **Entities** — Database models and ORM definitions
 - **Framework Detection** — React components, Angular services, etc.
 
-### Token Savings
+### Project Exploration & Context
 
-CortexCode dramatically reduces the tokens needed to give AI assistants useful context:
+Use CortexCode to inspect structure and answer practical questions about a codebase:
 
-```bash
-$ cortexcode stats
-
-╭──────── Token Savings Analysis ────────╮
-│ Source files         154 files          │
-│ Raw project tokens   203,847            │
-│ Full index tokens    45,291             │
-│ Context query tokens 487                │
-│                                         │
-│ Tokens saved         203,360            │
-│ Savings              99.8%              │
-│ Compression ratio    418.6x             │
-╰─────────────────────────────────────────╯
-```
+- **Context lookup** — Pull relevant symbols and files for a query
+- **Call graph tracing** — Follow how behavior moves across modules
+- **Architecture visibility** — Inspect routes, entities, dependencies, and layers
+- **Reusable index** — Keep a structured project map that tools can build on
 
 ### Interactive HTML Documentation
 
@@ -158,7 +137,6 @@ cortexcode wiki --no-modules       # Skip per-module pages (faster)
 - **Mermaid diagrams** — Auto-generated flow diagrams
 - **Concept mapping** — Maps technical concepts to symbols and files
 - **Concept search** — Ask "how does authentication work?" and get grounded answers
-- **Token tracking** — See exactly how many tokens each page used
 
 **Output:** `.cortexcode/wiki/index.html` — Open directly or serve locally.
 
@@ -209,12 +187,12 @@ CortexCode supports both a short form (`cc`) and full form (`cortexcode`). Comma
 | Command | Description |
 |---------|-------------|
 | `cc analyze context [query]` | Get relevant context for AI |
-| `cc analyze context [query] --tokens` | Show token savings for query |
+| `cc analyze context [query] --tokens` | Show context size estimates for a query |
 | `cc analyze search [query]` | Grep-like symbol search with type/file filters |
 | `cc analyze find [query]` | Semantic search by meaning ("auth handler") |
 | `cc analyze diff` | Show changed symbols since last commit |
 | `cc analyze diff --ref HEAD~3` | Compare against any git ref |
-| `cc analyze stats` | Show project stats and token savings |
+| `cc analyze stats` | Show project index and analysis stats |
 | `cc analyze scan` | Scan dependencies for security warnings |
 | `cc analyze trace <symbol>` | Trace code flow through call graph |
 | `cc analyze flow <concept>` | Analyze code flow grouped by file |
@@ -319,7 +297,7 @@ Add to `~/.windsurf/config.json`:
 
 ```bash
 # Paste this output into your AI chat
-cortexcode context "useAuth" --tokens
+cortexcode context "useAuth"
 ```
 
 ### 2. JSON Index (programmatic)
@@ -516,7 +494,6 @@ CortexCode respects `.gitignore` files (including nested ones) and has built-in
 ## Roadmap
 
 - [x] MCP server for direct AI agent integration
-- [x] Tiktoken-based accurate token counting
 - [x] Semantic search over symbols (TF-IDF + synonym expansion)
 - [x] Cross-file type inference
 - [x] Git diff-aware context (show only changed symbols)

{cortexcode-0.9.0 → cortexcode-0.10.0}/cortexcode/ai_docs/doc_generator.py

@@ -60,7 +60,7 @@ class AIDocGenerator:
             messages = prompts.generate_project_overview_prompt(index_data)
             output.overview = self._generate(messages, "overview")
             if output.overview:
-                (output_dir / "AI_OVERVIEW.md").write_text(output.overview)
+                (output_dir / "AI_OVERVIEW.md").write_text(output.overview, encoding="utf-8")
                 print("  -> AI_OVERVIEW.md")
         
         if "api" in docs:
@@ -68,7 +68,7 @@ class AIDocGenerator:
             messages = prompts.generate_api_docs_prompt(index_data)
             output.api_docs = self._generate(messages, "api")
             if output.api_docs:
-                (output_dir / "AI_API.md").write_text(output.api_docs)
+                (output_dir / "AI_API.md").write_text(output.api_docs, encoding="utf-8")
                 print("  -> AI_API.md")
         
         if "architecture" in docs:
@@ -76,7 +76,7 @@ class AIDocGenerator:
             messages = prompts.generate_architecture_prompt(index_data)
             output.architecture = self._generate(messages, "architecture")
             if output.architecture:
-                (output_dir / "AI_ARCHITECTURE.md").write_text(output.architecture)
+                (output_dir / "AI_ARCHITECTURE.md").write_text(output.architecture, encoding="utf-8")
                 print("  -> AI_ARCHITECTURE.md")
         
         if "flows" in docs:
@@ -84,7 +84,7 @@ class AIDocGenerator:
             messages = prompts.generate_flow_docs_prompt(index_data)
             output.flows = self._generate(messages, "flows")
             if output.flows:
-                (output_dir / "AI_FLOWS.md").write_text(output.flows)
+                (output_dir / "AI_FLOWS.md").write_text(output.flows, encoding="utf-8")
                 print("  -> AI_FLOWS.md")
         
         return output
@@ -107,7 +107,7 @@ class AIDocGenerator:
         result = self._generate(messages, f"module_{module_name}")
         
         if output_path and result:
-            Path(output_path).write_text(result)
+            Path(output_path).write_text(result, encoding="utf-8")
         
         return result
 

{cortexcode-0.9.0 → cortexcode-0.10.0}/cortexcode/diagrams/architecture.py

@@ -35,6 +35,21 @@ def generate_architecture_diagram(index_data: dict[str, Any]) -> str:
                 lines.append(f"        {framework_id}[\"{framework.get('name', 'unknown')} ({framework.get('count', 0)})\"]")
             lines.append("    end")
 
+        # Nx project graph overlay
+        nx_graph = project_profile.get("nx_project_graph", {})
+        nx_projects = project_profile.get("nx_projects", [])
+        if nx_graph and nx_projects:
+            lines.append("    subgraph nx_workspace [Nx Workspace]")
+            for proj_name in nx_projects:
+                proj_id = sanitize_id(f"nx_{proj_name}")
+                lines.append(f"        {proj_id}[\"{proj_name}\"]")
+            lines.append("    end")
+            for proj_name, deps in nx_graph.items():
+                proj_id = sanitize_id(f"nx_{proj_name}")
+                for dep in deps[:10]:
+                    dep_id = sanitize_id(f"nx_{dep}")
+                    lines.append(f"        {proj_id} --> {dep_id}")
+
         return "\n".join(lines)
 
     files = index_data.get("files", {})

{cortexcode-0.9.0 → cortexcode-0.10.0}/cortexcode/indexer.py

@@ -27,6 +27,7 @@ from cortexcode.indexing.resolution import (
     build_file_dependencies,
     build_type_map,
 )
+from cortexcode.indexing.nx_projects import parse_nx_workspace, build_nx_project_graph
 from cortexcode.plugins import plugin_registry
 
 
@@ -223,18 +224,32 @@ class CodeIndexer(IndexerExtractorMixin):
     
     def _build_index(self, root: Path) -> dict[str, Any]:
         """Build the final index structure."""
+        nx_workspace = parse_nx_workspace(root)
+        tsconfig_paths = nx_workspace.get("tsconfig_paths", {}) if nx_workspace else {}
+
+        def _build_file_deps(ts_paths=None):
+            return build_file_dependencies(self.file_symbols, ts_paths)
+
+        def _build_type_map(ts_paths=None):
+            return build_type_map(self.file_symbols, ts_paths)
+
+        def _build_project_profile(_root, file_deps, nx=None):
+            return build_project_profile(self.file_symbols, self.call_graph, file_deps, nx)
+
         result = build_index_result(
             root=root,
             file_symbols=self.file_symbols,
             call_graph=self.call_graph,
             timestamp=timestamp_now(),
             file_hashes=compute_hashes(root, self.file_symbols),
-            build_file_dependencies_fn=lambda: build_file_dependencies(self.file_symbols),
-            build_type_map_fn=lambda: build_type_map(self.file_symbols),
-            build_project_profile_fn=lambda _root, file_deps: build_project_profile(self.file_symbols, self.call_graph, file_deps),
+            build_file_dependencies_fn=lambda ts_paths=tsconfig_paths: _build_file_deps(ts_paths),
+            build_type_map_fn=lambda ts_paths=tsconfig_paths: _build_type_map(ts_paths),
+            build_project_profile_fn=lambda _root, file_deps, nx=nx_workspace: _build_project_profile(_root, file_deps, nx),
             language_map=LANGUAGE_MAP,
             regex_languages=REGEX_LANGUAGES,
             plugin_registry=plugin_registry,
+            nx_workspace=nx_workspace,
+            tsconfig_paths=tsconfig_paths,
         )
         # Add source code for context retrieval
         result["source_code"] = self.source_code

{cortexcode-0.9.0 → cortexcode-0.10.0}/cortexcode/indexing/build.py

@@ -18,6 +18,8 @@ def build_index_result(
     language_map: LanguageMap,
     regex_languages: RegexLanguages,
     plugin_registry,
+    nx_workspace: dict[str, Any] | None = None,
+    tsconfig_paths: dict[str, str] | None = None,
 ) -> dict[str, Any]:
     languages = set()
     for file_path in file_symbols.keys():
@@ -30,9 +32,9 @@ def build_index_result(
         else:
             languages.add(ext.lstrip("."))
 
-    file_deps = build_file_dependencies_fn()
-    type_map = build_type_map_fn()
-    project_profile = build_project_profile_fn(root, file_deps)
+    file_deps = build_file_dependencies_fn(tsconfig_paths)
+    type_map = build_type_map_fn(tsconfig_paths)
+    project_profile = build_project_profile_fn(root, file_deps, nx_workspace)
 
     result = {
         "project_root": str(root),
@@ -48,4 +50,9 @@ def build_index_result(
     if type_map:
         result["type_map"] = type_map
 
+    if nx_workspace:
+        result["nx_workspace"] = nx_workspace
+        if tsconfig_paths:
+            result["nx_workspace"]["tsconfig_paths"] = tsconfig_paths
+
     return plugin_registry.run_post_processors(result)

{cortexcode-0.9.0 → cortexcode-0.10.0}/cortexcode/indexing/defaults.py

@@ -4,7 +4,7 @@ DEFAULT_IGNORE_PATTERNS = {
     "*.egg-info", ".eggs", "*.pyc", "*.pyo",
     "node_modules", ".npm", ".yarn", ".pnpm-store", "bower_components",
     "dist", "build", "out", "output", "target", "bin", "obj",
-    ".build", "_build", "public/build", "lib",
+    ".build", "_build", "public/build",
     ".next", ".nuxt", ".svelte-kit", ".angular", ".turbo",
     ".parcel-cache", ".webpack", ".rollup.cache", ".vite",
     ".expo", ".gradle", "Pods", "DerivedData",

{cortexcode-0.9.0 → cortexcode-0.10.0}/cortexcode/indexing/filtering.py

@@ -22,7 +22,7 @@ def should_ignore_file(
     rel_posix = rel_path.as_posix()
 
     for pattern in default_ignore_patterns:
-        if pattern in path_str or pattern in path_posix or pattern in rel_posix:
+        if pattern in rel_str or pattern in rel_posix:
             return True
 
     for pattern in exclude_patterns:

{cortexcode-0.9.0 → cortexcode-0.10.0}/cortexcode/indexing/imports_exports.py

@@ -139,6 +139,13 @@ def find_js_exports(node: Any, exports: list[dict[str, Any]], get_node_name: Get
                                 name = export_clause_child.child_by_field_name("name")
                                 if name:
                                     exports.append({"name": name.text.decode("utf-8"), "type": "named"})
+            elif child.type == "export_clause":
+                # Barrel re-export: export { Button } from './lib/button'
+                for export_clause_child in child.children:
+                    if export_clause_child.type == "export_specifier":
+                        name = export_clause_child.child_by_field_name("name")
+                        if name:
+                            exports.append({"name": name.text.decode("utf-8"), "type": "re-export"})
             elif child.type == "variable_declaration":
                 for declaration_child in child.children:
                     if declaration_child.type == "variable_declarator":