Decepticon: Role-boundary forgery via ChatML special-token literals in web crawl output composed into LLM context
Key facts
- Attack vector
- Network
- Privileges required
- None
- User interaction
- None
- Attack complexity
- Low
- Impact type
- —
- Confidentiality · integrity · availability
- High · High · High
- Fix status
- Fixed
Attack profile from CVSS 3.1 · CNA. Fix status from CNA and GitHub.
Summary
Decepticon is an autonomous hacking agent for red teams. Versions prior to 1.1.17 wrap web crawl results — the output of agent reconnaissance against target services — into LLM messages without neutralizing ChatML special-token literals. Under the BYOK (Bring Your Own Key) deployment model, users configure their own LLM credentials to any OpenAI-compatible endpoint. Most open-source and self-deployed model providers (vLLM, SGLang, Ollama, LM Studio, text-generation-webui, etc.) do not filter special-token literals from user content in their default configurations. Those literals are parsed into structural role-boundary token IDs, meaning an attacker string planted in a target web page forges a new operator turn the model treats as authoritative, bypassing Decepticon's agent guardrails and resulting in arbitrary command execution inside the Kali Linux sandbox. Version 1.1.17 patches the issue.
Why it matters now
SSVC decision points
CISA ADP · SSVC 2.0.3
- Exploitation
- Poc
- Automatable
- Yes
- Technical impact
- Total
The decision (Track, Attend, Act) also depends on your mission and well-being impact; see CISA's SSVC guide.
Affected products
| Vendor | Product | Affected versions | Fixed in | Status | Source |
|---|---|---|---|---|---|
| BitterSecurity | Decepticon | before 1.1.17 | 1.1.17 | Patch available | CNA |
| BitterSecurity | decepticon-core | before 1.1.17 | 1.1.17 | Patch available | CNA |
| BitterSecurity | decepticon-sdk | before 1.1.17 | 1.1.17 | Patch available | CNA |
| — | decepticon-core (PyPI) | before 1.1.17 | 1.1.17 | Patch available | GHSA-g5f9-3xfg-p9mf |
| — | decepticon (PyPI) | before 1.1.17 | 1.1.17 | Patch available | GHSA-g5f9-3xfg-p9mf |
| — | decepticon-sdk (PyPI) | before 1.1.17 | 1.1.17 | Patch available | GHSA-g5f9-3xfg-p9mf |
- Decepticon VendorBitterSecurity Affected before 1.1.17 Fixed in 1.1.17 Patch available CNA
- decepticon-core VendorBitterSecurity Affected before 1.1.17 Fixed in 1.1.17 Patch available CNA
- decepticon-sdk VendorBitterSecurity Affected before 1.1.17 Fixed in 1.1.17 Patch available CNA
- decepticon-core (PyPI) Vendor— Affected before 1.1.17 Fixed in 1.1.17 Patch available GHSA-g5f9-3xfg-p9mf
- decepticon (PyPI) Vendor— Affected before 1.1.17 Fixed in 1.1.17 Patch available GHSA-g5f9-3xfg-p9mf
- decepticon-sdk (PyPI) Vendor— Affected before 1.1.17 Fixed in 1.1.17 Patch available GHSA-g5f9-3xfg-p9mf
Remediation
- Fix · CNA Decepticon (BitterSecurity) Fixed in 1.1.17
- Fix · CNA decepticon-core (BitterSecurity) Fixed in 1.1.17
- Fix · CNA decepticon-sdk (BitterSecurity) Fixed in 1.1.17
- Fix · GHSA-g5f9-3xfg-p9mf decepticon-core (PyPI) Fixed in 1.1.17 Advisory →
- Fix · GHSA-g5f9-3xfg-p9mf decepticon (PyPI) Fixed in 1.1.17 Advisory →
- Fix · GHSA-g5f9-3xfg-p9mf decepticon-sdk (PyPI) Fixed in 1.1.17 Advisory →
Timeline
- 2026-10-02 Threatist record created from CVE-2026-61732
- 2026-10-03 Alias added: GHSA-g5f9-3xfg-p9mf
References
Other 5
- https://github.com/BitterSecurity/Decepticon/security/advisories/GHSA-g5f9-3xfg-p9mf
- https://github.com/BitterSecurity/Decepticon/commit/79ee2aaf22f4c36a5b1968f6ca3f8086b6e35b67
- https://github.com/BitterSecurity/Decepticon/pull/715
- https://github.com/BitterSecurity/Decepticon/releases/tag/v1.1.17
- https://github.com/PurpleAILAB/Decepticon
Sources
- CVE-2026-61732 · CVE Program
- CVE-2026-61732 · NVD
- FIRST EPSS
- GHSA-g5f9-3xfg-p9mf · GitHub Advisory Database · CC BY 4.0
This product uses the NVD API but is not endorsed or certified by the NVD. Sources and licences