llamafile out-of-bounds read in ZIP container parser
- Exploited
- No known exploitation
- Fix
- Fix or patch referenced
- Who is exposed
- Not reported
Summary
Attackers can exploit a flaw in mozilla-ai llamafile's container parser, specifically in the llamafile_open_zip function, to read memory outside the intended bounds. The bug can be reached remotely without logging in, but it requires a user to take some action.
Written by a model
Original description
A flaw has been found in mozilla-ai llamafile up to .0.10.5. This issue affects the function llamafile_open_zip of the file llamafile/llamafile.c of the component Container Parser. This manipulation of the argument size causes out-of-bounds read. The attack is possible to be carried out remotely. Upgrading to version 0.10.6 is capable of addressing this issue. Patch name: 5ac1c4dbb76737fee71a54922df95342c667ba6f. It is advisable to upgrade the affected component.
What to do
Apply the patch listed under Evidence.
Check your systems with an agent
Paste this into a coding or ops agent. It checks your systems read-only and reports a table per asset.
Check whether my systems are affected by CVE-2026-133498 (mozilla-ai llamafile Container llamafile.c llamafile_open_zip out-of-bounds), using the cve.now data below.
Rules: work read-only. Do not run exploit code, proof-of-concept scripts or scanners that send attack payloads, and do not change any configuration, package, image or deployment. Report findings only.
What cve.now has for CVE-2026-133498 (entry last updated 2026-11-20):
Affected:
- mozilla-ai llamafile: versions .0.10.1, .0.10.2, .0.10.3, .0.10.4, .0.10.5
Fixed versions: none listed yet.
Setup: unknown whether the bug needs a non-default feature or setting. Check the advisory.
CISA KEV: CVE-2026-133498 is not listed as of cve.now's last read.
This KEV status is for CVE-2026-133498 only. If you also check another CVE, look up its own KEV status; never assume it from this one.
How to check:
- Installed software: check what is actually running, not what a repo declares. Read the version from each host, container image and Kubernetes node (the binary's version output, the OS package database, image layers), and check whether the conditions above hold there. Distribution packages may carry a backported fix under an older version number; confirm against the distribution's own advisory.
Verify against the linked advisories before concluding, and cite the date you read each one:
- https://github.com/mozilla-ai/llamafile/pull/1044
- https://github.com/mozilla-ai/llamafile/commit/5ac1c4dbb76737fee71a54922df95342c667ba6f
- https://github.com/mozilla-ai/llamafile/releases/tag/0.10.6
- https://github.com/mozilla-ai/llamafile/issues/1038
- https://vuldb.com/vuln/440918
- https://vuldb.com/vuln/440918/cti
- https://vuldb.com/cve/CVE-2026-133498
- https://vuldb.com/submit/1050809
Do not conclude an asset is unaffected unless you checked it and the evidence covers both its version and the conditions above. If you could not check something, mark it unknown and say what is missing.
Report a table with one row per asset and these columns: Asset | Version or build | Status (affected, fixed, not applicable, unknown) | Compensating controls | Exposure (internet-facing, internal, none) | Evidence | Next action.
Entry: https://cve.now/CVE-2026-133498
For follow-up questions, use the cve.now MCP server at https://mcp.cve.now/mcp (no authentication), for example execute with name cve.get and arguments {"id": "CVE-2026-133498"}.Affected versions
- mozilla-ai llamafile.0.10.1, .0.10.2, .0.10.3, .0.10.4, .0.10.5
cve.now Index
- Exploitation0/40No signal
- Exposure10/25CISA rates it automatable · Unknown: No advisory context yet
- Reach?/20Unknown. No tracked technology or package matched
- Severity5.3/10CVSS 5.3
- No patch0/5No signal
What would change the score
- Would score about 70 if added to CISA KEV
- Would score about 65 if CISA's SSVC rates exploitation active
- Would score about 30 if advisories confirm default setups are exposed
Timeline
- CVE published
- First seen by cve.now
- CISA SSVC assessment
Exploitation: none · Automatable: yes · Technical impact: partial
- Record last updated
Evidence 11
| Kind | Source |
|---|---|
| CVE record 2Establishes: The identifier and description. | |
| Record | CVE List · CVE-2026-133498 |
| Record | NVD · CVE-2026-133498 |
| Other links 9 | |
| Advisory | VulDB · Advisory #440918 |
| Advisory | VulDB · Threat intelligence #440918 |
| Advisory | VulDB · CVE-2026-133498 record |
| Advisory | VulDB · Submission #1050809 |
| Patch | GitHub · mozilla-ai/llamafile · Pull request #1044 |
| Patch | GitHub · mozilla-ai/llamafile · Commit 5ac1c4db |
| Patch | GitHub · mozilla-ai/llamafile · Release 0.10.6 |
| Report | GitHub · mozilla-ai/llamafile · Issue #1038 |
| Repository | GitHub · mozilla-ai/llamafile · Repository |
Questions about CVE-2026-133498
Is CVE-2026-133498 being exploited?
It is not in CISA's Known Exploited Vulnerabilities catalog, and CISA's SSVC assessment does not mark exploitation as active. That does not mean it has not been attacked: vendors and researchers can report attacks before CISA lists them. No public exploit code is tracked.
Source: cve.now feed data
Which versions are affected by CVE-2026-133498?
llamafile: versions .0.10.1, .0.10.2, .0.10.3, .0.10.4, .0.10.5.
Source: cve.now feed data
Related
- CVE-2026-77347Pozeiden: Out-of-bounds read/write (crash / memory corruption) on valid diagrams in grid-layout renderers
- CVE-2026-62321Netatalk: afpd Spotlight RPC heap-buffer-overflow in nested sl_pack_filemeta() marshalling
- CVE-2026-49388Netatalk: Heap out-of-bounds read in afpd Spotlight RPC unmarshalling via unbounded toc_index in sl_unpack_loop()
- CVE-2026-49387Netatalk: Heap out-of-bounds read in afpd Spotlight RPC unmarshalling via unbounded element count in sl_unpack_ints() and sibling handlers
- CVE-2026-45356Netatalk: Unsigned Integer Underflow in Spotlight RPC Count Decrement (sl_unpack_loop)
- CVE-2026-45355Netatalk: Integer Underflow in Spotlight RPC String Length Calculation (sl_unpack_cpx)
No corrections so far. Something wrong on this page? Report an error.