HOL LogoGuard

Explore HOL

  • HOL home
  • AI agent registry
  • AI plugins
  • Open standards
  • HOL members
  • Sponsor HOL
  • Advertise with HOL

Guard product

  • Guard overviewLocal security and control for AI agents and the tools they use.
  • FeaturesRuntime protection, policy routing, review, and evidence.

Explore Guard

  • Product previewWalk through Guard surfaces in read-only demo mode.
  • ComparisonCompare Guard with native controls and AI security vendors.

AI tools

  • All AI toolsEvery supported AI tool and how Guard applies policy to it.
  • Codex
  • Claude Code
  • Cursor
  • Antigravity CLI
  • OpenCode
  • Hermes
  • OpenClaw
  • GitHub Copilot CLI
  • Antigravity
  • Kimi
  • Grok
  • Pi / Oh My Pi
  • Zcode

Extensions

  • All extensionsBrowse command and MCP coverage with owners and stated limits.
  • Command coverageShell command protection across clouds, databases, backups, and packages.
  • MCP server coverageSee how Guard maps risk state across MCP tools and servers.
  • Core safetyThe safety floor listings that ship with Guard.
  • Data and resilienceBackup and storage command protection.
  • Cloud and infrastructureAWS, Azure, GCP, Kubernetes, and more.

Security

  • AI security hubSecurity research, advisories, and agent safety coverage.
  • AI tool securitySecurity profiles for each supported coding agent.
  • Safe labsHands-on attack simulations with safe boundaries.
  • Redacted warningsReal blocked actions with sensitive details removed.
  • AdvisoriesCoordinated disclosure reports for AI tooling.
  • Active CVEsSearch active CVEs affecting AI tooling.

Learn

  • Security guidesPractical guides for securing AI agent workflows.
  • DocsInstall, configure, and operate Guard with confidence.
  • ResearchPublished security research, benchmarks, and methodology.

Community

  • ReleasesVersion history, shipped changes and upgrade notes.
  • ContributorsThe people and contributions behind HOL Guard.
  • AffiliatesShare Guard with your audience and earn from referrals.
  • SponsorFund agreed research, implementation and shared testing through the HOL Partner Program.
  • AdvertiseCompare README logo placements, previews and monthly prices, or promote a plugin or extension.
PricingEnterpriseOpen AppInstall Guard
  1. Guard
  2. Security
  3. CVEs
  4. CVE 2026 29777 traefik kubernetes gateway rule injection via
HOL Guard

Public security guidance for teams protecting AI harnesses, MCP servers, skills, prompts, and local tool execution.

Install Guard

AI Security

  • Prompt injection
  • MCP security
  • OWASP MCP mapping
  • Supply chain

Resources

  • Trust packet
  • Harness setup
  • Redacted warnings
  • Safe labs

Product

  • Install Guard
  • Pricing
  • Open dashboard
Guard
  • Guard Overview
  • Releases
  • Contributors
  • Install Guard
  • Pricing
Docs
  • Documentation Index
  • Developer Hub
  • API Reference
  • Root OpenAPI
  • Registry OpenAPI
  • Run in Postman
  • Standards
  • Submit ERC-8004 Contract
  • Feature Your Agent
Best Plugins
  • Browse Plugins
  • Plugin Launches
  • Best Claude Plugins
  • Best Codex Plugins
  • Best Grok Plugins
  • Best Kimi Plugins
  • Best DeepSeek Plugins
  • Best Antigravity Plugins
  • Best MCP Servers
  • Best Cursor Plugins
  • Best OpenCode Plugins
Best Agents
  • Best ERC-8004 Agents
  • Best Virtuals Agents
  • Best MCP Servers
  • Best A2A Agents
  • Best x402 Payable
  • All Categories
Community
  • Telegram
  • X
More
  • About HOL
  • Sponsor
  • Advertise
  • Contact
  • Blog
  • GitHub
  • Privacy
  • Terms of Service
Settings

Copyright © 2026 HOL DAO LLC. All rights reserved.

Back to active CVEs
Unknown severity · Source severity not reportedCVE-2026-29777GHSA-8Q2W-WR49-WHQJ

Traefik: kubernetes gateway rule injection via unescaped backticks in HTTPRoute match valuesCVE-2026-29777

Answer in brief

CVE-2026-29777 records a Unknown severity vulnerability in Traefik: kubernetes gateway rule injection via unescaped backticks in HTTPRoute match values. The current sources do not mark it as known exploited. The current feed maps github.com/traefik/traefik (go), github.com/traefik/traefik/v2 (go), github.com/traefik/traefik/v3 (go). Check affected ranges and fixed versions before updating.

Analysis pending evidence review

HOL Guard separates source facts from reviewed analysis. See the methodology.

Published Mar 11, 2026Updated Sep 10, 2026Source checked Oct 25, 2026First seen by HOL Sep 10, 2026Material review Sep 10, 2026
Upstream Advisory

Key facts

Risk
Unknown severitySource severity not reported
Exploitation
Not marked as known exploited
Affected software
3 mapped packages or products
Fix availability
Available

Why this deserves its current priority

A CVSS score is not reported in the current record. The current sources do not mark it as known exploited. Treat this as a source-backed prioritization signal, not a statement about your environment.

Analysis status

Analysis pending evidence review

Factual feed record only; HOL analysis is not approved for indexing. Read the methodology.

Affected scope and exposure questions

The current feed maps github.com/traefik/traefik (go), github.com/traefik/traefik/v2 (go), github.com/traefik/traefik/v3 (go). Check affected ranges and fixed versions before updating.

Mapped affected packages and fixed versions
PackageAffected rangeFixed version
github.com/traefik/traefikgo>=0Not reported
github.com/traefik/traefik/v2go>=0Not reported
github.com/traefik/traefik/v3go>=0 <3.6.103.6.10

Recommended response

  1. 1Check inventory. Check lockfiles and deployed manifests for github.com/traefik/traefik, github.com/traefik/traefik/v2, github.com/traefik/traefik/v3.
  2. 2Review the reported fix. Update github.com/traefik/traefik/v3 to 3.6.10 if you use the affected versions. Test the change in a non-production environment first.

Evidence timeline and material changes

  1. Published upstream

    Mar 11, 2026

    Evidence: source:osv:source_dates:source-dates:record
  2. Source modified

    Sep 10, 2026

    Evidence: source:osv:source_dates:source-dates:record
  3. First seen by HOL

    Sep 10, 2026

Sources and claim methodology

  • GitHub security advisorygithub.com
  • GitHub security advisorygithub.com
  • GitHub security advisorygithub.com
  • NVD vulnerability recordnvd.nist.gov
Upstream source description

## Summary There is a potential vulnerability in Traefik's Kubernetes Gateway provider related to rule injection. A tenant with write access to an HTTPRoute resource can inject backtick-delimited rule tokens into Traefik's router rule language via unsanitized header or query parameter match values. In shared gateway deployments, this can bypass listener hostname constraints and redirect traffic for victim hostnames to attacker-controlled backends. ## Patches - https://github.com/traefik/traefik/releases/tag/v3.6.10 ## For more information If you have any questions or comments about this advisory, please [open an issue](https://github.com/traefik/traefik/issues). <details> <summary>Original Description</summary> hey Traefik, repo: https://github.com/traefik/traefik commit: a4a91344edcdd6276c1b766ca19ee3f0e346480f (as-of 2026-03-02) traefik's kubernetes gateway provider builds router rules by interpolating HTTPRoute match values into the traefik rule language using backtick-delimited string literals (e.g., `Header(`name`,`value`)`, `Query(`name`,`value`)`) without escaping or validation. because backtick is a delimiter in the rule language, a tenant-controlled backtick can terminate the literal and inject additional rule tokens (for example `) || HostRegexp(`.\*`) || ...`). this changes the parsed ast so that an injected OR branch is not gated by the intended `Host(...)` constraint due to operator precedence, and can result in end-to-end routing hijack (victim host routed to attacker backends). in shared gateway deployments that rely on gateway API listener hostname constraints to isolate tenants, this can enable cross-tenant routing hijack to attacker-controlled backends. ## expected vs actual expected: provider-generated rules must be injection-safe; tenant-controlled match values must not be able to change the rule parse tree beyond literal argument content, especially across listener hostname-constraint boundaries in shared gateway deployments. actual: a backtick inside a header/query match value can inject an OR branch into the generated rule, changing the ast root from `and` to `or` and enabling hostname-constraint bypass. ## severity HIGH (impact ceiling may reach the top severity tier in shared gateway threat models; end-to-end kubernetes reproduction is recommended to demonstrate cross-tenant routing impact). CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:C/C:H/I:H/A:N = 8.7 cwe: CWE-74 (improper neutralization of special elements in output used by a downstream component) ## affected versions - confirmed vulnerable at: a4a91344edcdd6276c1b766ca19ee3f0e346480f (pinned commit) - release matrix: not yet confirmed (needs version mapping for gateway api provider in v3) ## affected code - `pkg/provider/kubernetes/gateway/httproute.go`: `buildHeaderRules` and `buildQueryParamRules` build `Header(`%s`,`%s`)` / `Query(`%s`,`%s`)` without escaping - `pkg/provider/kubernetes/gateway/grpcroute.go`: `buildGRPCHeaderRules` builds `Header(`%s`,`%s`)` / `HeaderRegexp(`%s`,`%s`)` without escaping - `pkg/provider/kubernetes/knative/kubernetes.go`: `buildRule` builds `Header(`%s`,`%s`)` without escaping - the generated rule string is parsed by `pkg/muxer/http/parser.go` (predicate-based rule parser) - github permalinks (pinned): - https://github.com/traefik/traefik/blob/a4a91344edcdd6276c1b766ca19ee3f0e346480f/pkg/provider/kubernetes/gateway/httproute.go#L742 - https://github.com/traefik/traefik/blob/a4a91344edcdd6276c1b766ca19ee3f0e346480f/pkg/provider/kubernetes/gateway/httproute.go#L761 ## root cause the kubernetes gateway provider formats rule strings using backticks as string delimiters: ```go rules = append(rules, fmt.Sprintf("Header(`%s`,`%s`)", header.Name, header.Value)) rules = append(rules, fmt.Sprintf("Query(`%s`,`%s`)", qp.Name, qp.Value)) ``` if `header.Value` (or `qp.Value`) contains a backtick and operator tokens, it can terminate the literal and inject additional rule-language tokens, changing the parse tree. ## attacker control attacker-controlled input is the kubernetes control plane object `HTTPRoute` in a tenant namespace. the attacker controls: 1. `HTTPRoute.Spec.Rules[].Matches[].Headers[].Value` and/or `QueryParams[].Value` (string) 2. the payload content, including backticks and rule tokens ## impact in shared gateway setups, this can bypass gateway API listener hostname constraints, causing requests for victim hostnames to be routed to attacker backends. downstream effects can include credential/token capture and request forgery, depending on the workload behind the gateway. traefik's documentation frames gateway API as providing safer multi-tenant primitives via listener constraints (see https://doc.traefik.io/traefik/security/multi-tenant-kubernetes/). rule injection breaks those constraints when they are relied upon as a boundary. ## reproduction (attachment: poc.zip) attachment includes `poc.zip` with an integration PoC that: - shows canonical behavior where injection changes the parsed ast root to `or` and routes `victim.com` to the attacker handler (emits `[PROOF_MARKER]`) - shows a negative control using injection-safe quoting (`%q`) where the ast root remains `and` and routes `victim.com` to the victim handler (emits `[NC_MARKER]`) run canonical: ```bash unzip poc.zip -d poc cd poc make canonical ``` canonical output excerpt: ``` [CALLSITE_HIT] [PROOF_MARKER] ``` run control: ```bash unzip poc.zip -d poc cd poc make control ``` control output excerpt: ``` [NC_MARKER] ``` ## recommended fix encode rule arguments using injection-safe quoting (for example `fmt.Sprintf("Header(%q,%q)", name, value)`), or otherwise reject/escape backticks and other rule-language metacharacters before interpolation. add regression tests that include backticks and operator tokens inside header/query match values and assert they cannot change the parse tree. **fix accepted when:** tenant-controlled HTTPRoute match values cannot inject operators into the generated rule string and cannot change the resulting parsed ast structure. [[poc.zip](https://github.com/user-attachments/files/25698814/poc.zip)](https://github.com/user-attachments/files/25698814/poc.zip) [[PR_DESCRIPTION.md](https://github.com/user-attachments/files/25698815/PR_DESCRIPTION.md)](https://github.com/user-attachments/files/25698815/PR_DESCRIPTION.md) [[attack_scenario.md](https://github.com/user-attachments/files/25698816/attack_scenario.md)](https://github.com/user-attachments/files/25698816/attack_scenario.md) cheers, Oleh Konko </details>

Quoted source text, attributed separately from HOL analysis.

Record context

Vulnerability class
Secret Exfiltration
EPSS
Not reported
CWE IDs
Not reported
Source
OSV
Source checked
Oct 25, 2026
References
4 linked sources
Open source record