Vulnerability advisory

CVE-2026-113431

Unbounded decoded HTTP/2 header-list size in cowlib cow_http2_machine enables HPACK amplification memory exhaustion

Published Updated OSV record ↗

Also known as EEF-CVE-2026-113431

Severity
  • CVSS 4.0CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N

Affected versions

Repository

Affected commitsFixed by commit
GIT github.com/ninenines/cowlib ↗
b461b119e78e4e09bb28b186b09da7ed4a86a0dd to before a3b15e29b3067644bfd580246c87f7af7a90b3b8a3b15e29b3067644bfd580246c87f7af7a90b3b8
Not specifiedb461b119e78e4e09bb28b186b09da7ed4a86a0dd

Description

Allocation of Resources Without Limits or Throttling vulnerability in ninenines cowlib allows an unauthenticated remote peer to exhaust the memory of an HTTP/2 endpoint via HPACK-indexed header fields.

cow_http2_machine:headers_decode/4 in src/cow_http2_machine.erl decodes every HEADERS block with cow_hpack:decode/3. It bounds the compressed block size and the number of fields, but not the size of the decoded header list that RFC 9113 section 6.5.2 places under SETTINGS_MAX_HEADER_LIST_SIZE. A peer stores one header value of up to the dynamic table size (4096 bytes by default) and then references it with one-byte indexed fields, so about 100 bytes on the wire decode to a header list of about 390 KB. A consumer that copies the decoded values, as Cowboy does when cowboy_http2:headers_to_map/2 joins repeated cookie fields into one binary, allocates that amount for every stream. The total across streams and connections is not bounded.

This issue affects cowlib: from 2.7.0 onward.

References

  • Codex
  • Claude Code
  • VS Code
  • GitHub Copilot
  • Cursor
  • Cline
  • Windsurf
  • Google Anti-gravity
  • OpenCode
  • Pi
  • Gemini CLI
  • Hermes Agent
  • Zed
  • Junie
  • Qwen Code
  • Kiro
  • Kilo Code
  • Factory Droid

Set up GitHits for your agents

Run this command:

<> bash
npx githits@latest init

What happens next:

  1. Opens your browser to authorize GitHits
  2. Asks whether to set up GitHits at the user or project level
  3. Detects your supported AI coding tools
  4. Lets you choose which tools to configure
  5. Configures the selected tools