CVE-2026-61732
Decepticon: Role-boundary forgery via ChatML special-token literals in web crawl output composed into LLM context
Decepticon wraps 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.
- No known exploitation
- Severity
- Critical, CVSS 10.0 (v3.1)
- EPSS
- 1.22% chance of exploitation in 30 days (EPSS, as of Oct 23, 2026)
- Affected packages
- decepticon-core (pip) <= 1.1.16, fixed in 1.1.17 · GHSA-g5f9-3xfg-p9mf
- decepticon (pip) <= 1.1.16, fixed in 1.1.17 · GHSA-g5f9-3xfg-p9mf
- decepticon-sdk (pip) <= 1.1.16, fixed in 1.1.17 · GHSA-g5f9-3xfg-p9mf