← Back to Feed CACHED · 2026-09-10 11:29:32 · CACHE_KEY CVE-2026-82533
CVE-2026-82533 · CWE-807 · Disclosed 2026-09-08

DeepSeek Harness before 0.1.2-alpha.1 contains an authentication bypass vulnerability in its local HTTP…

ASSESSED — NOISGATE V0.5
Vendor
Reassessed
Verdict:
Do you agree?
01 · The Real Story

Someone left a skeleton key under the doormat, but the door is inside a locked house that almost nobody lives in

CVE-2026-82533 is an authentication bypass in DeepSeek Harness, a Python client library (pip install deepseek-harness) for DeepSeek V4-Pro/V4-Flash models. Versions before 0.1.2-alpha.1 expose a local HTTP control-plane API that validates the client-supplied Host header instead of the actual TCP connection origin (CWE-807). An attacker who can trick a user into visiting a malicious page while the harness is running locally can spoof the Host header via DNS rebinding or similar browser-mediated attacks, bypassing authentication entirely. This grants privileged command invocation, policy escalation to unrestricted execution, access to stored conversation histories, and full agent control. The fix is in commit 3e24087.

The vendor rates this CRITICAL at 9.6 — and the auth bypass mechanism is genuinely poor engineering. But the severity label wildly overstates the real-world risk. This is an alpha-stage developer tool with ~15,000 total PyPI downloads, listening on localhost only, requiring user interaction (the victim must click a link while the harness process is active). The blast radius caps at a single developer's AI agent session and chat history. There is no lateral movement, no credential harvesting, no domain-wide impact. The CVSS vector's S:C (Changed scope) is defensible in theory — the attacker crosses from browser context to local agent — but the practical population at risk is vanishingly small. This is a textbook case of CVSS inflation on a niche tool.

"Localhost-only AI client lib with 15K downloads — real bug, inflated blast radius."
02 · The Attack Path

4 steps from start to impact.

STEP 01

Victim runs DeepSeek Harness locally

The developer starts the DeepSeek Harness process, which opens a local HTTP control-plane API on a localhost port. This is the prerequisite — the API must be actively listening. Given the alpha status and 15K download count, the fraction of machines running this at any given time is extremely small.
Conditions required:
  • DeepSeek Harness < 0.1.2-alpha.1 installed and running
  • Local HTTP control-plane API is listening
Where this breaks in practice:
  • Only ~15K total PyPI installs; active concurrent instances are a tiny fraction of that
  • Alpha-stage software — few production or persistent deployments
  • Process must be running at the moment of attack
STEP 02

Attacker delivers malicious page to victim's browser

The attacker hosts a page containing JavaScript that performs DNS rebinding or a crafted fetch/XHR request targeting the harness's localhost port with a spoofed Host header. The victim must navigate to this page (UI:R). This is the browser-mediated pivot from remote to local.
Conditions required:
  • Victim visits attacker-controlled or compromised web page
  • Victim's browser must not block the rebinding or localhost request
Where this breaks in practice:
  • Requires social engineering or watering-hole to get the click
  • Modern browsers increasingly restrict localhost access from remote origins
  • DNS rebinding requires attacker-controlled DNS infrastructure and timing
Detection/coverage: Browser-level DNS rebinding protections; network monitoring for anomalous DNS TTL behavior
STEP 03

Host header spoofing bypasses authentication

The harness API checks only the Host header value to determine if the request is legitimate, rather than validating the TCP source address. The attacker's rebinding payload sends requests with a trusted Host value (e.g., 127.0.0.1 or localhost), passing the check. No credentials are needed.
Conditions required:
  • API validates Host header only (CWE-807)
  • DNS rebinding or CORS misconfiguration allows the spoofed request to land
Where this breaks in practice:
  • Some local firewalls or endpoint agents may block unexpected localhost connections from browser context
STEP 04

Privileged agent control achieved

With authentication bypassed, the attacker can invoke privileged commands, escalate session policies to unrestricted execution, read all stored conversation histories, and take full control of the AI agent. Impact is confined to the local harness instance — there is no documented lateral movement capability or credential exposure beyond the agent's own session.
Conditions required:
  • Successful Host header bypass from step 3
Where this breaks in practice:
  • Blast radius is a single user's AI agent session and chat logs
  • No pivot to network, AD, or other infrastructure
  • Data value depends entirely on what the developer discussed with the model
Detection/coverage: Application-level logging of API commands if enabled in harness config
03 · Intelligence Metadata

The supporting signals.

In-the-Wild ExploitationNo known exploitation. Not listed in CISA KEV. No campaigns or threat actor usage reported as of 2026-09-10.
Proof of ConceptNone public. No PoC repositories found on GitHub or Exploit-DB. The attack requires DNS rebinding infrastructure, raising the bar for casual reproduction.
EPSS0.00416 (0.416%) — 35th percentile. Well below the threshold that typically indicates near-term exploitation.
KEV StatusNot listed. No CISA KEV entry as of 2026-09-10.
CVSS VectorCVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H — Vendor score 9.6. The AV:N is technically accurate (browser-mediated) but misleading since the target API is localhost-only. UI:R correctly reflects required user interaction.
Affected VersionsDeepSeek Harness < 0.1.2-alpha.1 (all prior alpha releases)
Fixed Version0.1.2-alpha.1 (commit 3e24087)
Installed Base~15,000 total PyPI downloads per pepy.tech. Version 0.2.0 accounts for 78% of installs. Extremely niche; concurrent running instances likely number in the low hundreds at most.
Disclosure Date2026-09-08 — 2 days ago
Reporter / AdvisoryGHSA-8m2g-8cgm-3vcp. Maintainer: Henry Zhang (CyberWizard). No credited external researcher.
04 · The Call

noisgate verdict.

Final Verdict
DOWNGRADED to MEDIUM (5.0/10)

The single most decisive factor is the vanishingly small installed base and localhost-only attack surface — ~15K total PyPI downloads of an alpha-stage developer tool, requiring the victim to be actively running the process and to click a malicious link simultaneously. No high-value deployment role applies because DeepSeek Harness is a client-side AI chat wrapper, not infrastructure software.

HIGH Vulnerability mechanism (Host header bypass is well-documented in advisory)
HIGH Installed base size (PyPI download data is public)
MEDIUM Absence of in-the-wild exploitation (limited visibility into private threat intel)

Why this verdict

  • Localhost-only API: The control-plane listens on 127.0.0.1. Remote exploitation requires a browser-mediated pivot (DNS rebinding), adding significant friction and dependency on user interaction. This is not a directly internet-reachable service.
  • Tiny installed base: ~15,000 total PyPI downloads for an alpha-stage library. The concurrent running population is likely in the low hundreds. Fewer than 0.001% of enterprise hosts will have this installed, let alone running.
  • User interaction required: The CVSS vector correctly includes UI:R. The victim must visit an attacker-controlled page while the harness is actively running — a narrow timing window on a rare application.
  • Role multiplier: DeepSeek Harness is a Python client library for AI model interaction. It is not deployed in any high-value infrastructure role (not an IdP, hypervisor, CI/CD component, backup agent, kernel-mode agent, network edge, or database). (a) *Low-value role (developer workstation)*: canonical deployment — chain succeeds, blast radius is one user's chat history and local agent session. (b) *Typical role*: same as (a) — there is no server-mode deployment. (c) *High-value role*: no realistic high-value deployment exists. Even if used in a CI/CD pipeline, it wraps AI chat calls and does not manage build artifacts, secrets, or deployment credentials. No floor override applies.
  • No exploitation evidence: Zero PoCs, no KEV listing, EPSS at 35th percentile. The DNS rebinding prerequisite raises the bar above opportunistic scanning.

Why not higher?

Upgrading to HIGH would require either a meaningful high-value-role deployment pattern, active exploitation, or a directly internet-reachable attack surface. None of these conditions are met. The localhost-only listener, browser-mediated pivot, user interaction requirement, and ~15K-download alpha tool profile all compound to severely narrow the reachable population and blast radius.

Why not lower?

The underlying bug is a genuine authentication bypass (not just information disclosure), and the impact upon successful exploitation — unrestricted agent control and full conversation history access — is non-trivial for affected users. A developer's conversation history with an AI model could contain proprietary code, internal architecture details, or credentials pasted into prompts. LOW would understate the confidentiality impact for the small population actually at risk.

05 · Compensating Control

What to do — in priority order.

  1. Stop running DeepSeek Harness if below 0.1.2-alpha.1 — The simplest mitigation: pip install --upgrade deepseek-harness>=0.1.2a1. Since this is a developer tool with no uptime SLA, upgrading or stopping the process eliminates the attack surface entirely. No noisgate mitigation SLA applies for MEDIUM — go straight to the 365-day remediation window, though the trivial fix warrants immediate action.
  2. Bind the control-plane API to a Unix socket instead of TCP if configurable — If the harness supports Unix domain socket configuration, switching from TCP localhost to a socket eliminates the DNS rebinding vector entirely since browsers cannot target Unix sockets.
  3. Use browser DNS rebinding protections — Ensure developer workstations use DNS resolvers that enforce minimum TTLs and block rebinding patterns (e.g., Cloudflare Gateway, NextDNS, or enterprise DNS with rebinding protection enabled).
What doesn't work
  • WAF / reverse proxy in front of the API — the API listens on localhost and is not behind any network infrastructure; a WAF cannot intercept loopback traffic.
  • Network segmentation / firewall rules — the attack originates from the victim's own browser on the same host; network-level controls between hosts are irrelevant to a localhost-to-localhost attack path.
06 · Verification

Crowdsourced verification payload.

Run this on any developer workstation where DeepSeek Harness may be installed. No elevated privileges required. Invoke with: bash check_cve_2026_82533.sh

noisgate-verify.sh
BASHREAD-ONLYSAFE
#!/usr/bin/env bash
# CVE-2026-82533 checker for DeepSeek Harness
# Checks installed version against the fixed version 0.1.2-alpha.1
# Exit codes: 0 = PATCHED, 1 = VULNERABLE, 2 = UNKNOWN

set -euo pipefail

PKG="deepseek-harness"
FIXED="0.1.2a1"  # PEP 440 normalized form of 0.1.2-alpha.1

# Check if package is installed
if ! pip show "$PKG" &>/dev/null; then
  echo "UNKNOWN — $PKG is not installed on this system."
  exit 2
fi

INSTALLED=$(pip show "$PKG" 2>/dev/null | grep -i '^Version:' | awk '{print $2}')

if [ -z "$INSTALLED" ]; then
  echo "UNKNOWN — could not determine installed version of $PKG."
  exit 2
fi

# Use Python to compare PEP 440 versions reliably
RESULT=$(python3 -c "
from packaging.version import Version
import sys
try:
    installed = Version('$INSTALLED')
    fixed = Version('$FIXED')
    if installed >= fixed:
        print('PATCHED')
    else:
        print('VULNERABLE')
except Exception as e:
    print(f'UNKNOWN — version parse error: {e}')
    sys.exit(2)
")

echo "$RESULT — $PKG version $INSTALLED installed (fixed in >= $FIXED)"

case "$RESULT" in
  PATCHED)    exit 0 ;;
  VULNERABLE) exit 1 ;;
  *)          exit 2 ;;
esac
07 · Bottom Line

If you remember one thing.

TL;DR
This is a downgraded MEDIUM. DeepSeek Harness is an alpha-stage Python client library with ~15K total downloads — if it's in your environment at all, it's on a handful of developer workstations. The auth bypass is real but requires DNS rebinding plus user interaction to exploit a localhost-only API, and no exploitation has been observed. Per the noisgate remediation SLA for MEDIUM, no mitigation SLA applies — go straight to the 365-day remediation window. That said, the fix is a one-line pip install --upgrade, so there's no reason to wait: have any developers running it upgrade to ≥0.1.2-alpha.1 at their next convenience. Monday morning, send a Slack message to your ML/AI platform channel asking anyone using deepseek-harness to run pip install --upgrade deepseek-harness. Confirm with the verification script. Do not burn a change-control window on this.

Sources

  1. GHSA-8m2g-8cgm-3vcp — GitHub Security Advisory
  2. NVD — CVE-2026-82533
  3. PyPI — deepseek-harness
  4. pepy.tech — deepseek-harness download stats
  5. CWE-807 — Reliance on Untrusted Inputs in a Security Decision
  6. CISA KEV Catalog
Peer Review

What defenders are saying.

Submit a review attribution: handle + country only
0 flags selected · stored anonymously
Validation Results

Crowdsourced verification outputs.

Results submitted by users who ran the verification payload against their environment.