mycode-ai 0.1.0__tar.gz

mycode_ai-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Machine Adjacent Systems
+
+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.

mycode_ai-0.1.0/PKG-INFO

@@ -0,0 +1,161 @@
+Metadata-Version: 2.4
+Name: mycode-ai
+Version: 0.1.0
+Summary: Stress-testing tool for AI-generated code
+Author: Machine Adjacent Systems
+License-Expression: MIT
+Project-URL: Homepage, https://github.com/Manav-Guha/mycode
+Project-URL: Repository, https://github.com/Manav-Guha/mycode
+Project-URL: Issues, https://github.com/Manav-Guha/mycode/issues
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Software Development :: Testing
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Dynamic: license-file
+
+# myCode
+
+**Built it with AI? Test it before it breaks.**
+
+Stress-testing tool for AI-generated code. Point it at your project, describe what it does, and find out where it breaks under real-world conditions.
+
+## What It Does
+
+You built an app with ChatGPT, Cursor, Claude, or Copilot. It works on your machine. But what happens when 100 users hit it at once? When the CSV grows to 100,000 rows? When someone uploads a 50MB file? When the API you depend on goes slow?
+
+myCode answers these questions. It reads your code, understands your dependencies, generates stress scenarios, runs your actual code under escalating conditions, and tells you — in plain language — where it breaks.
+
+## Quick Start
+
+```bash
+pip install mycode-ai
+mycode ~/path/to/your/project
+```
+
+That's it. myCode will:
+1. Ask you what your project does (in your words, not engineering language)
+2. Show you what it plans to test and let you approve
+3. Run stress tests in a safe sandbox (your original files are never touched)
+4. Give you a diagnostic report
+
+## What It Tests
+
+- **Data volume scaling** — what happens as your data grows from 100 rows to 100,000
+- **Memory pressure** — does your app leak memory over repeated use
+- **Concurrent load** — multiple users or processes hitting shared resources
+- **Edge case inputs** — malformed data, empty values, unexpected types
+- **Dependency interactions** — where one component's failure cascades into another
+
+## Supported Languages
+
+- Python (Flask, FastAPI, Streamlit, pandas, NumPy, SQLAlchemy, LangChain, and more)
+- JavaScript/Node.js (React, Next.js, Express, Three.js, Prisma, Socket.io, and more)
+
+36 dependency profiles included. Unrecognized dependencies get generic stress testing.
+
+## Example Output
+
+```
+============================================================
+  myCode Diagnostic Report
+============================================================
+
+We found some problems that could affect your Incident Tracker under real-world conditions.
+
+- When your Incident Tracker is handling multiple users at once, requests timed out under load.
+- When your Incident Tracker is downloading large responses, the system hits memory limits.
+- When your Incident Tracker receives unexpected input, the code crashes instead of handling it gracefully.
+
+Found 3 critical issue(s) and 25 warning(s) across 70 scenarios.
+Performance degradation detected in 11 area(s).
+```
+
+## Offline Mode
+
+Run without any API calls:
+
+```bash
+mycode ~/path/to/your/project --offline
+```
+
+Uses template-based scenario generation instead of LLM. Still runs all stress tests and produces a full report.
+
+## Online Mode (Gemini Flash)
+
+With an API key, myCode uses Gemini Flash to generate a natural-language summary of findings:
+
+```bash
+export GEMINI_API_KEY=your_key_here
+mycode ~/path/to/your/project
+```
+
+Or pass it directly:
+
+```bash
+mycode ~/path/to/your/project --api-key your_key_here
+```
+
+Free tier: get a Gemini API key at https://aistudio.google.com/apikey (1,000 requests/day, no credit card).
+
+## How It Works
+
+1. **Ingester** parses your code (AST analysis), maps dependencies, identifies coupling points
+2. **Conversational interface** asks what your project does and how you expect it to be used
+3. **Scenario generator** creates stress test configurations based on your code + your intent
+4. **Execution engine** runs your actual code in a sandbox with synthetic data under escalating conditions
+5. **Report generator** translates raw results into plain-language diagnostics
+
+Your original files are never modified. Everything runs in a temporary environment that is destroyed on completion.
+
+## Safety
+
+- Your code is copied into a temporary sandbox — originals are never touched
+- Resource caps (memory, CPU, timeout) prevent runaway processes
+- Sandbox is destroyed on completion, crash, or interrupt (Ctrl+C)
+- No data is collected without explicit opt-in consent
+- No source code is ever sent to any API — only project descriptions and dependency names
+
+**Important:** myCode executes your project's code on your machine. Do not point it at code you do not trust.
+
+## What myCode Is NOT
+
+- Not a linter or code reviewer
+- Not a security scanner
+- Not a code generator — myCode never modifies your code
+- Not a guarantee of code quality — it's a diagnostic tool
+
+## Requirements
+
+- Python 3.10+
+- Node.js 18+ (for JavaScript project testing)
+
+## Validated Against Real Projects
+
+myCode has been tested against real-world projects across Python and JavaScript:
+
+| Project | Language | Scenarios | Critical Issues | Result |
+|---------|----------|-----------|-----------------|--------|
+| Multi-framework web app (Flask + FastAPI + Streamlit) | Python | 70 | 3 | Timeouts under load, memory scaling issues |
+| Analysis tool (React + charting libraries) | JavaScript | 25 | 4 | Edge case input crashes, memory growth |
+| Multi-component framework (React) | JavaScript | 19 | 0 | All clean |
+| Expense tracker (React) | JavaScript | 5 | 0 | Mostly clean |
+| CLI tool — myCode self-test (26 files, 24K lines) | Python | 279 | 0 | All clean |
+
+398 total scenarios, 7 critical issues found, 0 myCode errors.
+
+[See full anonymized test reports →](docs/validation-reports.md)
+
+## Contact
+
+Machine.Adjacent.Systems@protonmail.com
+
+## License
+
+MIT
+
+---
+
+*myCode diagnoses — it does not prescribe. Interpret results in your context.*

mycode_ai-0.1.0/README.md

@@ -0,0 +1,143 @@
+# myCode
+
+**Built it with AI? Test it before it breaks.**
+
+Stress-testing tool for AI-generated code. Point it at your project, describe what it does, and find out where it breaks under real-world conditions.
+
+## What It Does
+
+You built an app with ChatGPT, Cursor, Claude, or Copilot. It works on your machine. But what happens when 100 users hit it at once? When the CSV grows to 100,000 rows? When someone uploads a 50MB file? When the API you depend on goes slow?
+
+myCode answers these questions. It reads your code, understands your dependencies, generates stress scenarios, runs your actual code under escalating conditions, and tells you — in plain language — where it breaks.
+
+## Quick Start
+
+```bash
+pip install mycode-ai
+mycode ~/path/to/your/project
+```
+
+That's it. myCode will:
+1. Ask you what your project does (in your words, not engineering language)
+2. Show you what it plans to test and let you approve
+3. Run stress tests in a safe sandbox (your original files are never touched)
+4. Give you a diagnostic report
+
+## What It Tests
+
+- **Data volume scaling** — what happens as your data grows from 100 rows to 100,000
+- **Memory pressure** — does your app leak memory over repeated use
+- **Concurrent load** — multiple users or processes hitting shared resources
+- **Edge case inputs** — malformed data, empty values, unexpected types
+- **Dependency interactions** — where one component's failure cascades into another
+
+## Supported Languages
+
+- Python (Flask, FastAPI, Streamlit, pandas, NumPy, SQLAlchemy, LangChain, and more)
+- JavaScript/Node.js (React, Next.js, Express, Three.js, Prisma, Socket.io, and more)
+
+36 dependency profiles included. Unrecognized dependencies get generic stress testing.
+
+## Example Output
+
+```
+============================================================
+  myCode Diagnostic Report
+============================================================
+
+We found some problems that could affect your Incident Tracker under real-world conditions.
+
+- When your Incident Tracker is handling multiple users at once, requests timed out under load.
+- When your Incident Tracker is downloading large responses, the system hits memory limits.
+- When your Incident Tracker receives unexpected input, the code crashes instead of handling it gracefully.
+
+Found 3 critical issue(s) and 25 warning(s) across 70 scenarios.
+Performance degradation detected in 11 area(s).
+```
+
+## Offline Mode
+
+Run without any API calls:
+
+```bash
+mycode ~/path/to/your/project --offline
+```
+
+Uses template-based scenario generation instead of LLM. Still runs all stress tests and produces a full report.
+
+## Online Mode (Gemini Flash)
+
+With an API key, myCode uses Gemini Flash to generate a natural-language summary of findings:
+
+```bash
+export GEMINI_API_KEY=your_key_here
+mycode ~/path/to/your/project
+```
+
+Or pass it directly:
+
+```bash
+mycode ~/path/to/your/project --api-key your_key_here
+```
+
+Free tier: get a Gemini API key at https://aistudio.google.com/apikey (1,000 requests/day, no credit card).
+
+## How It Works
+
+1. **Ingester** parses your code (AST analysis), maps dependencies, identifies coupling points
+2. **Conversational interface** asks what your project does and how you expect it to be used
+3. **Scenario generator** creates stress test configurations based on your code + your intent
+4. **Execution engine** runs your actual code in a sandbox with synthetic data under escalating conditions
+5. **Report generator** translates raw results into plain-language diagnostics
+
+Your original files are never modified. Everything runs in a temporary environment that is destroyed on completion.
+
+## Safety
+
+- Your code is copied into a temporary sandbox — originals are never touched
+- Resource caps (memory, CPU, timeout) prevent runaway processes
+- Sandbox is destroyed on completion, crash, or interrupt (Ctrl+C)
+- No data is collected without explicit opt-in consent
+- No source code is ever sent to any API — only project descriptions and dependency names
+
+**Important:** myCode executes your project's code on your machine. Do not point it at code you do not trust.
+
+## What myCode Is NOT
+
+- Not a linter or code reviewer
+- Not a security scanner
+- Not a code generator — myCode never modifies your code
+- Not a guarantee of code quality — it's a diagnostic tool
+
+## Requirements
+
+- Python 3.10+
+- Node.js 18+ (for JavaScript project testing)
+
+## Validated Against Real Projects
+
+myCode has been tested against real-world projects across Python and JavaScript:
+
+| Project | Language | Scenarios | Critical Issues | Result |
+|---------|----------|-----------|-----------------|--------|
+| Multi-framework web app (Flask + FastAPI + Streamlit) | Python | 70 | 3 | Timeouts under load, memory scaling issues |
+| Analysis tool (React + charting libraries) | JavaScript | 25 | 4 | Edge case input crashes, memory growth |
+| Multi-component framework (React) | JavaScript | 19 | 0 | All clean |
+| Expense tracker (React) | JavaScript | 5 | 0 | Mostly clean |
+| CLI tool — myCode self-test (26 files, 24K lines) | Python | 279 | 0 | All clean |
+
+398 total scenarios, 7 critical issues found, 0 myCode errors.
+
+[See full anonymized test reports →](docs/validation-reports.md)
+
+## Contact
+
+Machine.Adjacent.Systems@protonmail.com
+
+## License
+
+MIT
+
+---
+
+*myCode diagnoses — it does not prescribe. Interpret results in your context.*

mycode_ai-0.1.0/pyproject.toml

@@ -0,0 +1,39 @@
+[build-system]
+requires = ["setuptools>=68.0"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "mycode-ai"
+version = "0.1.0"
+description = "Stress-testing tool for AI-generated code"
+readme = "README.md"
+requires-python = ">=3.10"
+authors = [
+    {name = "Machine Adjacent Systems"},
+]
+license = "MIT"
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Developers",
+    "Programming Language :: Python :: 3",
+    "Topic :: Software Development :: Testing",
+]
+
+[project.urls]
+Homepage = "https://github.com/Manav-Guha/mycode"
+Repository = "https://github.com/Manav-Guha/mycode"
+Issues = "https://github.com/Manav-Guha/mycode/issues"
+
+[project.scripts]
+mycode = "mycode.cli:main"
+
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[tool.setuptools.package-data]
+mycode = ["profiles/**/*.json"]
+
+[tool.pytest.ini_options]
+markers = [
+    "slow: integration tests that create real virtual environments",
+]

mycode_ai-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

mycode_ai-0.1.0/src/mycode/__init__.py

@@ -0,0 +1,145 @@
+"""myCode — Stress-testing tool for AI-generated code."""
+
+from mycode.ingester import (
+    CouplingPoint,
+    DependencyInfo,
+    FileAnalysis,
+    FunctionFlow,
+    IngestionError,
+    IngestionResult,
+    ProjectIngester,
+)
+from mycode.js_ingester import JsProjectIngester
+from mycode.library import (
+    ComponentLibrary,
+    DependencyProfile,
+    LibraryError,
+    ProfileMatch,
+    ProfileNotFoundError,
+)
+from mycode.scenario import (
+    LLMConfig,
+    LLMError,
+    LLMResponseError,
+    ScenarioError,
+    ScenarioGenerator,
+    ScenarioGeneratorResult,
+    StressTestScenario,
+)
+from mycode.engine import (
+    EngineError,
+    ExecutionEngine,
+    ExecutionEngineResult,
+    ScenarioResult,
+    StepResult,
+)
+from mycode.interface import (
+    ConversationalInterface,
+    InterfaceError,
+    InterfaceResult,
+    OperationalIntent,
+    ScenarioReview,
+    TerminalIO,
+    UserIO,
+    summarize_ingestion,
+)
+from mycode.report import (
+    DegradationPoint,
+    DiagnosticReport,
+    Finding,
+    ReportError,
+    ReportGenerator,
+)
+from mycode.recorder import (
+    ConsentError,
+    InteractionRecorder,
+    RecorderError,
+    SessionRecord,
+)
+from mycode.pipeline import (
+    LanguageDetectionError,
+    PipelineConfig,
+    PipelineError,
+    PipelineResult,
+    StageResult,
+    detect_language,
+    run_pipeline,
+)
+from mycode.session import (
+    DependencyInstallError,
+    EnvironmentInfo,
+    ResourceCaps,
+    SessionError,
+    SessionManager,
+    SessionResult,
+    VenvCreationError,
+)
+
+__all__ = [
+    # Session Manager (C1)
+    "SessionManager",
+    "ResourceCaps",
+    "EnvironmentInfo",
+    "SessionResult",
+    "SessionError",
+    "VenvCreationError",
+    "DependencyInstallError",
+    # Project Ingester (C2)
+    "ProjectIngester",
+    "IngestionResult",
+    "IngestionError",
+    "FileAnalysis",
+    "DependencyInfo",
+    "FunctionFlow",
+    "CouplingPoint",
+    # JS Project Ingester (C3)
+    "JsProjectIngester",
+    # Component Library (C4)
+    "ComponentLibrary",
+    "DependencyProfile",
+    "ProfileMatch",
+    "LibraryError",
+    "ProfileNotFoundError",
+    # Scenario Generator (D1)
+    "ScenarioGenerator",
+    "ScenarioGeneratorResult",
+    "StressTestScenario",
+    "LLMConfig",
+    "ScenarioError",
+    "LLMError",
+    "LLMResponseError",
+    # Execution Engine (D2)
+    "ExecutionEngine",
+    "ExecutionEngineResult",
+    "ScenarioResult",
+    "StepResult",
+    "EngineError",
+    # Pipeline (D3)
+    "run_pipeline",
+    "detect_language",
+    "PipelineConfig",
+    "PipelineResult",
+    "StageResult",
+    "PipelineError",
+    "LanguageDetectionError",
+    # Conversational Interface (E1)
+    "ConversationalInterface",
+    "InterfaceResult",
+    "OperationalIntent",
+    "ScenarioReview",
+    "InterfaceError",
+    "TerminalIO",
+    "UserIO",
+    "summarize_ingestion",
+    # Report Generator (E2)
+    "ReportGenerator",
+    "DiagnosticReport",
+    "Finding",
+    "DegradationPoint",
+    "ReportError",
+    # Interaction Recorder (E3)
+    "InteractionRecorder",
+    "SessionRecord",
+    "RecorderError",
+    "ConsentError",
+]

mycode_ai-0.1.0/src/mycode/__main__.py

@@ -0,0 +1,7 @@
+"""Allow running myCode as: python -m mycode"""
+
+import sys
+
+from mycode.cli import main
+
+sys.exit(main())

mycode_ai-0.1.0/src/mycode/cli.py

@@ -0,0 +1,177 @@
+"""myCode CLI — stress-test AI-generated code from the command line.
+
+Usage::
+
+    python -m mycode /path/to/project [options]
+
+Options::
+
+    --offline             Use offline mode (no LLM API calls)
+    --language LANG       Force language (python or javascript)
+    --consent             Opt in to anonymous interaction recording
+    --api-key KEY         API key for LLM backend
+    --api-base URL        Base URL for OpenAI-compatible API
+    --model MODEL         Model identifier
+    --skip-version-check  Skip PyPI/npm version lookups
+    --non-interactive     Skip conversational interface
+    --verbose / -v        Enable verbose logging
+"""
+
+import argparse
+import logging
+import os
+import sys
+from pathlib import Path
+
+from mycode.interface import TerminalIO
+from mycode.pipeline import PipelineConfig, run_pipeline
+from mycode.scenario import LLMConfig
+
+
+def build_parser() -> argparse.ArgumentParser:
+    """Build the CLI argument parser."""
+    parser = argparse.ArgumentParser(
+        prog="mycode",
+        description=(
+            "myCode — Stress-test AI-generated code before it breaks.\n\n"
+            "Point myCode at your project, describe what it does, "
+            "and get a plain-language report showing where it breaks."
+        ),
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+    )
+    parser.add_argument(
+        "project_path",
+        type=Path,
+        help="Path to the project directory to test",
+    )
+    parser.add_argument(
+        "--offline",
+        action="store_true",
+        default=False,
+        help="Run without LLM API calls (uses offline scenario generation)",
+    )
+    parser.add_argument(
+        "--language",
+        choices=["python", "javascript"],
+        default=None,
+        help="Force project language (auto-detected if not specified)",
+    )
+    parser.add_argument(
+        "--consent",
+        action="store_true",
+        default=False,
+        help="Opt in to anonymous interaction recording",
+    )
+    parser.add_argument(
+        "--api-key",
+        default=None,
+        help="API key for the LLM backend",
+    )
+    parser.add_argument(
+        "--api-base",
+        default=None,
+        help="Base URL for OpenAI-compatible API endpoint",
+    )
+    parser.add_argument(
+        "--model",
+        default=None,
+        help="LLM model identifier",
+    )
+    parser.add_argument(
+        "--skip-version-check",
+        action="store_true",
+        default=False,
+        help="Skip PyPI/npm version lookups (faster)",
+    )
+    parser.add_argument(
+        "--verbose", "-v",
+        action="store_true",
+        default=False,
+        help="Enable verbose logging",
+    )
+    parser.add_argument(
+        "--non-interactive",
+        action="store_true",
+        default=False,
+        help="Skip conversational interface (use default stress testing profile)",
+    )
+    return parser
+
+
+def main(argv: list[str] | None = None) -> int:
+    """CLI entry point.  Returns exit code (0=success, 1=failure)."""
+    parser = build_parser()
+    args = parser.parse_args(argv)
+
+    # Configure logging
+    level = logging.DEBUG if args.verbose else logging.WARNING
+    logging.basicConfig(level=level, format="%(levelname)s: %(message)s")
+
+    # Validate project path
+    project = Path(args.project_path).resolve()
+    if not project.exists():
+        print(f"Error: Project path does not exist: {project}", file=sys.stderr)
+        return 1
+    if not project.is_dir():
+        print(f"Error: Project path is not a directory: {project}", file=sys.stderr)
+        return 1
+
+    # Build LLM config — flag takes precedence, then env var
+    api_key = args.api_key or os.environ.get("GEMINI_API_KEY")
+    llm_config = None
+    if api_key:
+        kwargs: dict = {"api_key": api_key}
+        if args.api_base:
+            kwargs["base_url"] = args.api_base
+        if args.model:
+            kwargs["model"] = args.model
+        llm_config = LLMConfig(**kwargs)
+
+    # Offline mode: explicit flag or no API key available
+    offline = args.offline or (api_key is None)
+
+    # Build pipeline config
+    non_interactive = args.non_interactive or not sys.stdin.isatty()
+    config = PipelineConfig(
+        project_path=project,
+        language=args.language,
+        llm_config=llm_config,
+        offline=offline,
+        skip_version_check=args.skip_version_check,
+        consent=args.consent,
+        io=TerminalIO(),
+        operational_intent="General stress testing" if non_interactive else "",
+        auto_approve_scenarios=non_interactive,
+    )
+
+    # Run pipeline
+    result = run_pipeline(config)
+
+    # Display report
+    if result.report:
+        print(result.report.as_text())
+    elif result.execution:
+        print("\n--- Execution Results (report generation failed) ---")
+        print(f"Scenarios completed: {result.execution.scenarios_completed}")
+        print(f"Scenarios failed: {result.execution.scenarios_failed}")
+        if result.warnings:
+            print("\nWarnings:")
+            for w in result.warnings:
+                print(f"  - {w}")
+    else:
+        failed = result.failed_stage
+        if failed:
+            print(f"\nPipeline failed at stage: {failed}")
+        for s in result.stages:
+            if not s.success:
+                print(f"  Error ({s.stage}): {s.error}")
+        if result.warnings:
+            print("\nWarnings:")
+            for w in result.warnings:
+                print(f"  - {w}")
+
+    # Recording path
+    if result.recording_path:
+        print(f"\nSession recorded: {result.recording_path}")
+
+    return 0 if result.success else 1