webdev-pitfall-analyzer 2025.12.21133820

This package helps users identify and avoid common pitfalls in web and HTML development by analyzing text-based input—like code snippets, design descriptions, or technical discussions—and returning a

webdev-pitfall-analyzer

webdev-pitfall-analyzer helps developers quickly spot common web‑development pitfalls—security issues, accessibility problems, performance bottlenecks, outdated patterns, and more—by analysing short text inputs such as code snippets, design questions, or discussion fragments. The package uses an LLM (by default ChatLLM7 from langchain_llm7) to return a concise, structured list of warnings and best‑practice recommendations.

---

📦 Installation

pip install webdev_pitfall_analyzer

---

🚀 Quick start

from webdev_pitfall_analyzer import webdev_pitfall_analyzer

# Example HTML/JS snippet
snippet = """
<input type="text" onblur="doBadThing()" />
<script>
  function doBadThing() {
    eval(userInput);   // <- unsafe!
  }
</script>
"""

warnings = webdev_pitfall_analyzer(user_input=snippet)
print(warnings)

Output (example)

[
  "⚠️ Use of `eval` can lead to XSS vulnerabilities. Consider safer alternatives.",
  "⚠️ Missing `label` element for the input; this hurts accessibility.",
  "⚠️ Inline event handlers (`onblur`) are discouraged; attach listeners via JavaScript instead."
]

---

🛠️ Function signature

def webdev_pitfall_analyzer(
    user_input: str,
    api_key: Optional[str] = None,
    llm: Optional[BaseChatModel] = None,
) -> List[str]:

Parameter	Type	Description
user_input	str	The text (code, question, description, etc.) you want analysed.
api_key	Optional[str]	API key for ChatLLM7. If omitted, the function will look for the LLM7_API_KEY environment variable; if still not found it falls back to an unauthenticated request (may be rate‑limited).
llm	Optional[BaseChatModel]	Any LangChain‑compatible LLM instance. If not supplied, the default ChatLLM7 (from langchain_llm7) is used.

The function returns a list of warning/recommendation strings. If the underlying LLM call fails, a RuntimeError is raised.

---

🔧 Using a custom LLM

You can pass any LangChain chat model that follows the BaseChatModel interface.

OpenAI

from langchain_openai import ChatOpenAI
from webdev_pitfall_analyzer import webdev_pitfall_analyzer

my_llm = ChatOpenAI(model="gpt-4o-mini")
result = webdev_pitfall_analyzer(user_input="...", llm=my_llm)

Anthropic

from langchain_anthropic import ChatAnthropic
from webdev_pitfall_analyzer import webdev_pitfall_analyzer

my_llm = ChatAnthropic(model="claude-3-haiku-20240307")
result = webdev_pitfall_analyzer(user_input="...", llm=my_llm)

Google GenAI

from langchain_google_genai import ChatGoogleGenerativeAI
from webdev_pitfall_analyzer import webdev_pitfall_analyzer

my_llm = ChatGoogleGenerativeAI(model="gemini-1.5-flash")
result = webdev_pitfall_analyzer(user_input="...", llm=my_llm)

---

🔑 API key for ChatLLM7

The default free tier of LLM7 is sufficient for most occasional analyses.

• Set the key once in your environment:

export LLM7_API_KEY="your_llm7_key"

• Or pass it directly:

result = webdev_pitfall_analyzer(user_input="...", api_key="your_llm7_key")

Obtain a free key by registering at https://token.llm7.io/.

---

📚 What the analyzer looks for

• Security – unsafe functions (eval, innerHTML, etc.), injection risks.
• Accessibility – missing ARIA attributes, improper heading structure, unlabeled form controls.
• Performance – blocking synchronous XHR, large inline assets, unnecessary reflows.
• Modern best practices – deprecated APIs, outdated HTML5 elements, inefficient CSS selectors.

The underlying prompt (human_prompt, system_prompt) can be customised by forking the repository.

---

📂 Repository

🛠️ Issues & feature requests: https://github.com/chigwell/webdev_pitfall_analyzer/issues

---

👤 Author

Eugene Evstafev
Email: [e-mail]
GitHub: https://github.com/chigwell

---

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.