Skip to content

Security: LF3551/Open-IPv8-Lab

Security

SECURITY.md

Security Policy

Scope

Open-IPv8-Lab is an experimental, educational, userspace-only toolkit. It does not:

  • Handle production network traffic
  • Process real authentication credentials or sensitive data
  • Modify the kernel, network stack, or system configuration
  • Require root, raw sockets, or elevated privileges
  • Open external network connections (all demos use localhost)

In scope

  • Packet parser vulnerabilities (malformed input, buffer overflows in pure Python)
  • CLI injection via crafted arguments
  • Denial-of-service in simulation loops (infinite recursion, memory exhaustion)
  • Logic errors in security modules: ingress filtering, ACL8, prefix enforcement
  • Dependency vulnerabilities in typer, rich, textual, pyyaml

Out of scope

  • Kernel or OS-level network exploits (this is userspace only)
  • Vulnerabilities in the IPv8 specification itself (report to the IETF draft author)
  • Social engineering or phishing
  • Physical access attacks

Known limitations

  1. No real cryptography — mTLS and OAuth8 use mock implementations with placeholder keys. Do not use for production authentication.
  2. No input sanitization on YAML configs — pyyaml.safe_load() is used (no arbitrary code execution), but malicious YAML can cause excessive memory usage.
  3. Stateful CLI is process-scoped — commands like zone init store state in module-level variables, lost between process invocations. Not a security issue, but may surprise users.
  4. No rate limiting on CLI — the fuzzer and benchmark tools can consume significant CPU. This is by design for testing.
  5. WHOIS8/NetLog8 are mocks — they simulate protocol behaviour without real network I/O or persistent storage.

Fuzz testing

The built-in packet fuzzer (ipv8lab fuzz) tests protocol robustness with 9 mutation strategies:

Strategy Description
bit_flip Random bit flips in packet bytes
byte_random Replace random bytes with random values
boundary Insert boundary values (0x00, 0xFF, max int)
truncate Truncate packets to random lengths
extend Append random data beyond expected length
checksum Corrupt CRC32 checksums
field_mutate Mutate individual header fields
fragment Generate invalid fragmentation combinations
combined All strategies combined (default)

Running the fuzzer

# Default: 100 cases, combined strategy, parser target
ipv8lab fuzz run

# Heavy fuzz: 10,000 cases against all targets
ipv8lab fuzz run --count 10000 --target all

# Reproducible run with seed
ipv8lab fuzz run --count 1000 --seed 42 --json

# Specific strategy
ipv8lab fuzz run --strategy bit_flip --count 500

# List available strategies and targets
ipv8lab fuzz strategies
ipv8lab fuzz targets

Fuzz targets

Target What it tests
parser Packet deserialization — malformed headers, invalid fields
security Ingress filtering, prefix validation, ASN spoofing detection
fragment Fragmentation/reassembly — overlapping fragments, invalid offsets
routing Route table lookup — invalid prefixes, unreachable destinations
all All targets combined

Reporting a vulnerability

If you discover a security issue, please report it via email:

contact@alekseialeinikov.com

Please do not open a public GitHub issue for security vulnerabilities.

Include:

  • Description of the issue
  • Steps to reproduce (CLI command, input data, or Python snippet)
  • Expected vs actual behaviour
  • Impact assessment

Response

I will acknowledge receipt within 72 hours and provide a timeline for a fix.

There aren't any published security advisories