# Write Ida Script

> Write an IDAPython script using verified API workflows from the IDA SDK MCP server Use when this capability is needed.

- Skill: `tomevault-io/write-ida-script` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add tomevault-io/write-ida-script`
- Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/write-ida-script/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: tomevault-io (https://skillmd.com/u/tomevault-io)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/tomevault-io/write-ida-script

---


# Write IDA Script

Write an IDAPython script by first consulting the `ida-api-mcp` MCP tools to retrieve verified API call sequences, then composing the script from those patterns.

## Process

Follow these steps in order:

### 1. Decompose the request

Break the user's request into discrete sub-tasks. For example, "list all functions and their cross-references" becomes:
- Sub-task A: enumerate all functions
- Sub-task B: get cross-references for each function

### 2. Retrieve workflows

For each sub-task, call `get_workflows` with a natural-language description:

```
get_workflows("enumerate all functions in the database")
get_workflows("get cross references to a function")
```

### 3. Look up unfamiliar APIs

For any API function in the workflow results that you're not confident about, call `get_api_doc`:

```
get_api_doc("xrefblk_t")
get_api_doc("get_func_name")
```

### 4. Find companion APIs if needed

If a workflow seems incomplete (e.g., you have iteration but no formatting), call `list_related_apis`:

```
list_related_apis("get_func")
```

### 5. Write the script

Compose the script following these conventions:

- **Explicit imports**: `import ida_funcs`, not `from ida_funcs import *`
- **`main()` wrapper**: All logic inside `def main():` with `if __name__ == "__main__": main()`
- **None checks**: Always check return values — `get_func()`, `decompile()`, etc. can return `None`
- **`print()` for output**: Use `print()` in IDAPython, not `ida_kernwin.msg()`
- **Module-qualified calls**: `ida_funcs.get_func(ea)`, not bare `get_func(ea)`

Canonical style example:

```python
"""
Short summary of what the script does.

Longer description of the workflow: what it takes as input,
what it produces, and any prerequisites.

Usage: Run in IDA Pro via File -> Script file...
"""

import ida_funcs
import ida_hexrays
import ida_kernwin


def main():
    ea = ida_kernwin.ask_addr(0, "Enter function address:")
    if ea is None:
        return

    pfn = ida_funcs.get_func(ea)
    if pfn is None:
        print("No function found at 0x%X" % ea)
        return

    print("Function: 0x%X - 0x%X" % (pfn.start_ea, pfn.end_ea))

    cf = ida_hexrays.decompile(pfn.start_ea)
    if cf is None:
        print("Decompilation failed at 0x%X" % pfn.start_ea)
        return

    print(str(cf))


if __name__ == "__main__":
    main()
```

### 6. Explain the script

After writing the script, briefly explain:
- Which workflows/API patterns were used
- What each major section does
- Any limitations or assumptions

## Example

User: "write an IDAPython script that lists all functions with their sizes"

1. Sub-tasks: enumerate functions, compute size, format output
2. `get_workflows("enumerate all functions")` → reveals `idautils.Functions()` + `ida_funcs.get_func()`
3. `get_api_doc("get_func")` → confirms `func_t` has `.start_ea` and `.end_ea`
4. Write script using `idautils.Functions()` iterator, `ida_funcs.get_func()` for each, size = `end_ea - start_ea`

---
> Converted and distributed by [TomeVault](https://tomevault.io/claim/taardisaa) — claim your Tome and manage your conversions.
<!-- tomevault:4.0:skill_md:2026-04-16 -->