repointel 0.1.0__tar.gz

repointel-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Nirakar Jena
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

repointel-0.1.0/PKG-INFO

@@ -0,0 +1,316 @@
+Metadata-Version: 2.4
+Name: repointel
+Version: 0.1.0
+Summary: Local-first codebase-aware AI: index any repo and ask questions about it.
+Author-email: Nirakar Jena <jenashubham60@gmail.com>
+License-Expression: MIT
+Project-URL: Repository, https://github.com/nirakar24/reposcope
+Project-URL: Bug Tracker, https://github.com/nirakar24/reposcope/issues
+Keywords: codebase,rag,ai,search,developer-tools
+Classifier: Development Status :: 3 - Alpha
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Software Development :: Libraries
+Classifier: Topic :: Utilities
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Provides-Extra: api
+Requires-Dist: fastapi>=0.115; extra == "api"
+Requires-Dist: uvicorn>=0.30; extra == "api"
+Provides-Extra: openai
+Requires-Dist: openai>=1.0; extra == "openai"
+Provides-Extra: gemini
+Requires-Dist: google-genai>=0.6; extra == "gemini"
+Provides-Extra: claude
+Requires-Dist: anthropic>=0.40; extra == "claude"
+Provides-Extra: embed
+Requires-Dist: sentence-transformers>=3.0; extra == "embed"
+Requires-Dist: numpy>=1.26; extra == "embed"
+Provides-Extra: all
+Requires-Dist: fastapi>=0.115; extra == "all"
+Requires-Dist: uvicorn>=0.30; extra == "all"
+Requires-Dist: openai>=1.0; extra == "all"
+Requires-Dist: google-genai>=0.6; extra == "all"
+Requires-Dist: anthropic>=0.40; extra == "all"
+Requires-Dist: sentence-transformers>=3.0; extra == "all"
+Requires-Dist: numpy>=1.26; extra == "all"
+Provides-Extra: dev
+Requires-Dist: pytest>=8.0; extra == "dev"
+Dynamic: license-file
+
+# RepoScope
+
+A local-first tool for querying codebases in plain English. Point it at any repository, build an index, and ask questions — no cloud sync, no background services, no mandatory API keys.
+
+---
+
+## How it works
+
+**Without `--embed` (default):**
+```
+repo → file discovery → chunking → JSON index → BM25 retrieval → LLM answer
+```
+
+**With `--embed` (semantic search):**
+```
+repo → file discovery → chunking → JSON index + .npy embeddings → BM25 + vector → RRF merge → LLM answer
+```
+
+RepoScope walks your project and breaks files into chunks — by function, class, or method boundary for Python, JS/TS, and C# (using regex-based detection), and by fixed line windows for everything else. Chunks are stored in a local JSON index and ranked with BM25 scoring at query time. If you run `index --embed`, it also generates sentence embeddings and merges the two results with Reciprocal Rank Fusion for better semantic matches.
+
+If an LLM key is configured, `ask` feeds the top-ranked chunks to the model and returns a cited answer. If not, it falls back to showing the top matches with previews.
+
+---
+
+## Installation
+
+RepoScope is not yet on PyPI. Install from source:
+
+```bash
+git clone https://github.com/nirakar24/RepoScope.git
+cd RepoScope
+pip install -e .
+```
+
+To add optional features:
+
+```bash
+pip install -e ".[embed]"   # semantic search (downloads ~80 MB model on first use)
+pip install -e ".[claude]"  # Anthropic Claude for LLM answers
+pip install -e ".[gemini]"  # Google Gemini for LLM answers
+pip install -e ".[openai]"  # OpenAI for LLM answers and embeddings
+pip install -e ".[api]"     # FastAPI server
+pip install -e ".[all]"     # everything
+```
+
+---
+
+## Setting up an LLM API key
+
+`index` and `search` work with no API key. `ask` needs one to generate answers.
+
+### Interactive setup (recommended)
+
+```bash
+repointel configure
+```
+
+Prompts for your provider and key (input is hidden), then saves to `~/.config/repointel/.env`. Picked up automatically in every session and directory from then on.
+
+```
+Which provider do you want to use for 'ask' answers?
+  1) Anthropic Claude
+  2) Google Gemini
+  3) OpenAI
+
+Enter 1, 2, or 3: 2
+
+Google Gemini API key (input hidden):
+Saved GEMINI_API_KEY to ~/.config/repointel/.env
+```
+
+### Environment variable
+
+Add to your shell profile (`~/.bashrc`, `~/.zshrc`, etc.):
+
+```bash
+export ANTHROPIC_API_KEY="sk-ant-..."
+export GEMINI_API_KEY="AIza..."
+export OPENAI_API_KEY="sk-..."
+```
+
+### `.env` file
+
+Create a `.env` in your project directory or any parent. RepoScope walks up from the current directory automatically:
+
+```
+GEMINI_API_KEY=AIza...
+```
+
+### Provider priority and model overrides
+
+If multiple keys are present, the order is: **Claude → Gemini → OpenAI**.
+
+Override the default model with environment variables:
+
+```bash
+REPOSCOPE_CLAUDE_MODEL=claude-sonnet-4-6
+REPOSCOPE_GEMINI_MODEL=gemini-2.5-flash
+REPOSCOPE_OPENAI_MODEL=gpt-4.1-mini
+```
+
+### Where to get keys
+
+| Provider | Free tier | Key page |
+|---|---|---|
+| Google Gemini | Yes | https://aistudio.google.com/apikey |
+| Anthropic Claude | No | https://console.anthropic.com |
+| OpenAI | No | https://platform.openai.com/api-keys |
+
+---
+
+## Quick start
+
+```bash
+# Index your project
+repointel index /path/to/project
+
+# Search (instant, no LLM)
+repointel search "where is authentication handled"
+
+# Ask a question (requires an LLM key)
+repointel ask "how does the database schema relate to the API routes?"
+```
+
+---
+
+## Commands
+
+### `configure`
+Interactive first-time setup. Saves your LLM API key to `~/.config/repointel/.env`.
+
+```bash
+repointel configure
+```
+
+---
+
+### `index`
+Walks a directory, chunks its files, and writes a JSON index.
+
+```bash
+repointel index /path/to/project
+```
+
+Add `--embed` to generate sentence embeddings alongside the index. Once present, `search` and `ask` automatically switch to hybrid retrieval — no extra flag needed at query time.
+
+```bash
+repointel index /path/to/project --embed
+```
+
+Use `--index-file` (before the subcommand) to control where the index is written. Useful for keeping separate indexes per project:
+
+```bash
+repointel --index-file .reposcope/backend.json index ./backend --embed
+```
+
+The default path is `.reposcope/index.json` in the current directory.
+
+---
+
+### `search`
+Retrieves the most relevant chunks for a query. Instant — no network call.
+
+Uses BM25 by default. Automatically switches to hybrid BM25 + vector search if embeddings exist for the current index.
+
+```bash
+repointel search "JWT token validation"
+repointel search "database migration" --top-k 5
+repointel search "controller routes" --json
+```
+
+| Flag | Default | Description |
+|---|---|---|
+| `--top-k` | `8` | Number of results to return |
+| `--json` | off | Emit results as a JSON array |
+
+---
+
+### `ask`
+Retrieves top chunks and sends them to an LLM for a cited answer.
+
+```bash
+repointel ask "how does authentication work?"
+repointel ask "what entities exist in the database?" --top-k 12
+```
+
+Falls back to listing top matches with text previews if no LLM key is set.
+
+---
+
+### `stats`
+Prints a breakdown of the current index.
+
+```bash
+repointel stats
+```
+
+```json
+{
+  "files_indexed": 72,
+  "chunks_indexed": 428,
+  "languages": { "csharp": 160, "javascript": 25, "json": 221 },
+  "kinds": { "method": 117, "block": 247, "class": 25 },
+  "embeddings": ".reposcope/index.npy"
+}
+```
+
+---
+
+## Multiple projects
+
+Use `--index-file` to maintain separate indexes. The flag goes before the subcommand.
+
+```bash
+repointel --index-file .reposcope/frontend.json index ./frontend
+repointel --index-file .reposcope/backend.json  index ./backend
+
+repointel --index-file .reposcope/frontend.json ask "how is routing configured?"
+repointel --index-file .reposcope/backend.json  ask "what database tables exist?"
+```
+
+---
+
+## Optional REST API
+
+```bash
+pip install -e ".[api]"
+uvicorn reposcope.api:app --reload
+```
+
+| Method | Endpoint | Body |
+|---|---|---|
+| `POST` | `/index` | `{ "path": "/abs/path/to/repo", "embed": false }` |
+| `POST` | `/search` | `{ "query": "...", "top_k": 8 }` |
+| `POST` | `/ask` | `{ "query": "...", "top_k": 8 }` |
+| `GET` | `/stats` | — |
+
+Docs at `http://localhost:8000/docs`.
+
+---
+
+## Supported languages
+
+| Language | Chunking |
+|---|---|
+| Python | Regex-based: splits at `def` / `async def` / `class` boundaries |
+| JavaScript / TypeScript / JSX / TSX | Regex-based: splits at `function`, `class`, and arrow function boundaries |
+| C# | Regex-based: splits at class, interface, record, struct, and method boundaries |
+| SQL, Markdown, JSON, YAML, TOML, CSS, SCSS, HTML, Dockerfile, Makefile | Fixed 80-line windows with 15-line overlap |
+
+Files over 750 KB and generated lock files (`package-lock.json`, `yarn.lock`, `pnpm-lock.yaml`) are skipped.
+
+---
+
+## Ignored directories
+
+`node_modules`, `.git`, `dist`, `build`, `bin`, `obj`, `.venv`, `__pycache__`, `.next`, `.nuxt`, `coverage`, `target`, `temp`, `vendor`, and other standard build/cache directories are excluded automatically.
+
+---
+
+## Roadmap
+
+- [ ] Tree-sitter chunking for true AST-level boundaries (replacing the regex approach)
+- [ ] Incremental re-indexing on file change
+- [ ] Cross-repo index merging for monorepos
+- [ ] Qdrant / Chroma backend for repositories with >50k chunks
+
+---
+
+## License
+
+MIT

repointel-0.1.0/README.md

@@ -0,0 +1,272 @@
+# RepoScope
+
+A local-first tool for querying codebases in plain English. Point it at any repository, build an index, and ask questions — no cloud sync, no background services, no mandatory API keys.
+
+---
+
+## How it works
+
+**Without `--embed` (default):**
+```
+repo → file discovery → chunking → JSON index → BM25 retrieval → LLM answer
+```
+
+**With `--embed` (semantic search):**
+```
+repo → file discovery → chunking → JSON index + .npy embeddings → BM25 + vector → RRF merge → LLM answer
+```
+
+RepoScope walks your project and breaks files into chunks — by function, class, or method boundary for Python, JS/TS, and C# (using regex-based detection), and by fixed line windows for everything else. Chunks are stored in a local JSON index and ranked with BM25 scoring at query time. If you run `index --embed`, it also generates sentence embeddings and merges the two results with Reciprocal Rank Fusion for better semantic matches.
+
+If an LLM key is configured, `ask` feeds the top-ranked chunks to the model and returns a cited answer. If not, it falls back to showing the top matches with previews.
+
+---
+
+## Installation
+
+RepoScope is not yet on PyPI. Install from source:
+
+```bash
+git clone https://github.com/nirakar24/RepoScope.git
+cd RepoScope
+pip install -e .
+```
+
+To add optional features:
+
+```bash
+pip install -e ".[embed]"   # semantic search (downloads ~80 MB model on first use)
+pip install -e ".[claude]"  # Anthropic Claude for LLM answers
+pip install -e ".[gemini]"  # Google Gemini for LLM answers
+pip install -e ".[openai]"  # OpenAI for LLM answers and embeddings
+pip install -e ".[api]"     # FastAPI server
+pip install -e ".[all]"     # everything
+```
+
+---
+
+## Setting up an LLM API key
+
+`index` and `search` work with no API key. `ask` needs one to generate answers.
+
+### Interactive setup (recommended)
+
+```bash
+repointel configure
+```
+
+Prompts for your provider and key (input is hidden), then saves to `~/.config/repointel/.env`. Picked up automatically in every session and directory from then on.
+
+```
+Which provider do you want to use for 'ask' answers?
+  1) Anthropic Claude
+  2) Google Gemini
+  3) OpenAI
+
+Enter 1, 2, or 3: 2
+
+Google Gemini API key (input hidden):
+Saved GEMINI_API_KEY to ~/.config/repointel/.env
+```
+
+### Environment variable
+
+Add to your shell profile (`~/.bashrc`, `~/.zshrc`, etc.):
+
+```bash
+export ANTHROPIC_API_KEY="sk-ant-..."
+export GEMINI_API_KEY="AIza..."
+export OPENAI_API_KEY="sk-..."
+```
+
+### `.env` file
+
+Create a `.env` in your project directory or any parent. RepoScope walks up from the current directory automatically:
+
+```
+GEMINI_API_KEY=AIza...
+```
+
+### Provider priority and model overrides
+
+If multiple keys are present, the order is: **Claude → Gemini → OpenAI**.
+
+Override the default model with environment variables:
+
+```bash
+REPOSCOPE_CLAUDE_MODEL=claude-sonnet-4-6
+REPOSCOPE_GEMINI_MODEL=gemini-2.5-flash
+REPOSCOPE_OPENAI_MODEL=gpt-4.1-mini
+```
+
+### Where to get keys
+
+| Provider | Free tier | Key page |
+|---|---|---|
+| Google Gemini | Yes | https://aistudio.google.com/apikey |
+| Anthropic Claude | No | https://console.anthropic.com |
+| OpenAI | No | https://platform.openai.com/api-keys |
+
+---
+
+## Quick start
+
+```bash
+# Index your project
+repointel index /path/to/project
+
+# Search (instant, no LLM)
+repointel search "where is authentication handled"
+
+# Ask a question (requires an LLM key)
+repointel ask "how does the database schema relate to the API routes?"
+```
+
+---
+
+## Commands
+
+### `configure`
+Interactive first-time setup. Saves your LLM API key to `~/.config/repointel/.env`.
+
+```bash
+repointel configure
+```
+
+---
+
+### `index`
+Walks a directory, chunks its files, and writes a JSON index.
+
+```bash
+repointel index /path/to/project
+```
+
+Add `--embed` to generate sentence embeddings alongside the index. Once present, `search` and `ask` automatically switch to hybrid retrieval — no extra flag needed at query time.
+
+```bash
+repointel index /path/to/project --embed
+```
+
+Use `--index-file` (before the subcommand) to control where the index is written. Useful for keeping separate indexes per project:
+
+```bash
+repointel --index-file .reposcope/backend.json index ./backend --embed
+```
+
+The default path is `.reposcope/index.json` in the current directory.
+
+---
+
+### `search`
+Retrieves the most relevant chunks for a query. Instant — no network call.
+
+Uses BM25 by default. Automatically switches to hybrid BM25 + vector search if embeddings exist for the current index.
+
+```bash
+repointel search "JWT token validation"
+repointel search "database migration" --top-k 5
+repointel search "controller routes" --json
+```
+
+| Flag | Default | Description |
+|---|---|---|
+| `--top-k` | `8` | Number of results to return |
+| `--json` | off | Emit results as a JSON array |
+
+---
+
+### `ask`
+Retrieves top chunks and sends them to an LLM for a cited answer.
+
+```bash
+repointel ask "how does authentication work?"
+repointel ask "what entities exist in the database?" --top-k 12
+```
+
+Falls back to listing top matches with text previews if no LLM key is set.
+
+---
+
+### `stats`
+Prints a breakdown of the current index.
+
+```bash
+repointel stats
+```
+
+```json
+{
+  "files_indexed": 72,
+  "chunks_indexed": 428,
+  "languages": { "csharp": 160, "javascript": 25, "json": 221 },
+  "kinds": { "method": 117, "block": 247, "class": 25 },
+  "embeddings": ".reposcope/index.npy"
+}
+```
+
+---
+
+## Multiple projects
+
+Use `--index-file` to maintain separate indexes. The flag goes before the subcommand.
+
+```bash
+repointel --index-file .reposcope/frontend.json index ./frontend
+repointel --index-file .reposcope/backend.json  index ./backend
+
+repointel --index-file .reposcope/frontend.json ask "how is routing configured?"
+repointel --index-file .reposcope/backend.json  ask "what database tables exist?"
+```
+
+---
+
+## Optional REST API
+
+```bash
+pip install -e ".[api]"
+uvicorn reposcope.api:app --reload
+```
+
+| Method | Endpoint | Body |
+|---|---|---|
+| `POST` | `/index` | `{ "path": "/abs/path/to/repo", "embed": false }` |
+| `POST` | `/search` | `{ "query": "...", "top_k": 8 }` |
+| `POST` | `/ask` | `{ "query": "...", "top_k": 8 }` |
+| `GET` | `/stats` | — |
+
+Docs at `http://localhost:8000/docs`.
+
+---
+
+## Supported languages
+
+| Language | Chunking |
+|---|---|
+| Python | Regex-based: splits at `def` / `async def` / `class` boundaries |
+| JavaScript / TypeScript / JSX / TSX | Regex-based: splits at `function`, `class`, and arrow function boundaries |
+| C# | Regex-based: splits at class, interface, record, struct, and method boundaries |
+| SQL, Markdown, JSON, YAML, TOML, CSS, SCSS, HTML, Dockerfile, Makefile | Fixed 80-line windows with 15-line overlap |
+
+Files over 750 KB and generated lock files (`package-lock.json`, `yarn.lock`, `pnpm-lock.yaml`) are skipped.
+
+---
+
+## Ignored directories
+
+`node_modules`, `.git`, `dist`, `build`, `bin`, `obj`, `.venv`, `__pycache__`, `.next`, `.nuxt`, `coverage`, `target`, `temp`, `vendor`, and other standard build/cache directories are excluded automatically.
+
+---
+
+## Roadmap
+
+- [ ] Tree-sitter chunking for true AST-level boundaries (replacing the regex approach)
+- [ ] Incremental re-indexing on file change
+- [ ] Cross-repo index merging for monorepos
+- [ ] Qdrant / Chroma backend for repositories with >50k chunks
+
+---
+
+## License
+
+MIT

repointel-0.1.0/pyproject.toml

@@ -0,0 +1,41 @@
+[project]
+name = "repointel"
+version = "0.1.0"
+description = "Local-first codebase-aware AI: index any repo and ask questions about it."
+readme = "README.md"
+requires-python = ">=3.11"
+license = "MIT"
+license-files = ["LICENSE"]
+authors = [{name = "Nirakar Jena", email = "jenashubham60@gmail.com"}]
+keywords = ["codebase", "rag", "ai", "search", "developer-tools"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Environment :: Console",
+    "Intended Audience :: Developers",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Software Development :: Libraries",
+    "Topic :: Utilities",
+]
+dependencies = []
+
+[project.optional-dependencies]
+api    = ["fastapi>=0.115", "uvicorn>=0.30"]
+openai = ["openai>=1.0"]
+gemini = ["google-genai>=0.6"]
+claude = ["anthropic>=0.40"]
+embed  = ["sentence-transformers>=3.0", "numpy>=1.26"]
+all    = ["fastapi>=0.115", "uvicorn>=0.30", "openai>=1.0", "google-genai>=0.6", "anthropic>=0.40", "sentence-transformers>=3.0", "numpy>=1.26"]
+dev    = ["pytest>=8.0"]
+
+[project.scripts]
+repointel = "reposcope.cli:main"
+
+[project.urls]
+Repository = "https://github.com/nirakar24/reposcope"
+"Bug Tracker" = "https://github.com/nirakar24/reposcope/issues"
+
+[build-system]
+requires = ["setuptools>=68"]
+build-backend = "setuptools.build_meta"