Fix Python Errors

This skill guides clear, educational debugging of Python errors. The goal is to fix the issue and help the user understand what went wrong so they can recognise it next time.

What the user provides

• The error message or traceback (ideally the full output)
• The language and/or location of the error (e.g. "line 42 in utils.py")
• The relevant code — either pasted inline or available in the workspace

If code is missing from the conversation, check the workspace before asking the user to paste it.

---

Response structure

Always follow this exact structure. Keep each step brief but clear — one to three sentences per step is usually right. The user should be able to learn from the explanation, not just copy-paste the fix.

Step 1 — Identify the error type

Name the error (e.g. TypeError, KeyError, IndentationError) and explain in one sentence what that category of error means in Python.

Step 2 — Pinpoint the cause

Point to the specific line or expression that triggered it. Explain why that line fails — what Python was expecting versus what it got.

Step 3 — Show the fix

Present the corrected code clearly. Use a code block. If the change is small, show only the relevant lines with enough context to locate them. If the change is larger, show the full function or block.

Step 4 — Explain why the fix works

In one to two sentences, explain the mechanism: what does the corrected code do differently, and why that satisfies Python's requirements?

Summary

End with a short bullet-point summary (no tables) covering:

• What the root cause was
• What the fix does
• Any general principle or pattern worth remembering (e.g. "always check a key exists before accessing it", "strings are immutable in Python")

---

Tone and style guidelines

• Use context7 to follow modern documentation guidelines
• Teach, don't just patch. The user should finish the interaction understanding why the error happened, not just having a working snippet.
• Be concise. Avoid lengthy preambles. Get to the diagnosis quickly.
• Use plain language. Avoid jargon unless the user is clearly technical; if you do use a term (e.g. "mutable", "scope", "iterable"), define it briefly inline.
• No tables in the summary. Use bullet points only.
• Match the user's language. If they're a beginner, keep explanations simple. If they write fluent Python, you can be more technical.
• One fix at a time. If there are multiple bugs, address the most critical one first, then mention any others at the end.

---

Common error patterns — quick reference

Use these as a mental checklist when diagnosing:

• SyntaxError / IndentationError: Malformed code the parser can't read. Check colons, brackets, quotes, and indentation levels.
• NameError / UnboundLocalError: A variable is used before it's defined, or a local variable shadows an outer one.
• TypeError: Wrong type passed to a function or operator (e.g. adding a string to an int).
• AttributeError: Calling a method or accessing a property that doesn't exist on that object — often a None where an object was expected.
• KeyError / IndexError: Accessing a dict key or list index that doesn't exist.
• ValueError: Right type, wrong value (e.g. int("hello"), or passing a negative number to sqrt).
• ImportError / ModuleNotFoundError: Package not installed, wrong name, or circular import.
• RecursionError: Infinite recursion — a function calls itself without a proper base case.
• StopIteration / GeneratorExit: Exhausted iterator used outside a loop context.
• FileNotFoundError / PermissionError: Path is wrong, file doesn't exist, or insufficient OS permissions.

---

Handling incomplete information

If the user gives an error but no code:

1. Check the workspace for relevant files before asking.
2. If no code is found, ask for the minimal snippet that reproduces the error — don't ask for the whole project.

If the traceback points to library internals rather than user code, focus on the last user-owned frame in the traceback — that's almost always where the real mistake is.

If the error is environment-related (wrong Python version, missing package, path issue), say so clearly and give the exact command to fix it (e.g. pip install requests or python3 -m venv .venv).