Customers & proof
Signet governs a live agent fleet in production at Lavish Gains today, and its safety has been validated on public academic benchmarks TrajeXY did not write. Every benchmark figure below is recomputed from raw output — hover any underlined number for its plain-language explanation.
Customer · Lavish Gains
Lavish Gains is TrajeXY's flagship customer — a first-party personalization business running Signet across a fleet of AI agents in production. Every agent action is gated against a sealed mandate and written to a tamper-evident ledger, every day, in market.
Signet sits in the execution path of Lavish Gains' production agents — sending, updating, and acting on live systems under enforced mandates.
In production, Signet has blocked plaintext-secret leaks, unauthorized tier changes, and privilege-escalation attempts before they could take effect.
Every allow and deny is on a hash-chained ledger — a complete, checkable record of what each agent was permitted to do and why.
Independent validation
Beyond production, Signet has been tested on public agent-security benchmarks with a fixed model and Signet wired into the real code. The results below are recomputed per test from raw output. Customers and partners under review get the full multi-suite board, with every figure re-derived from raw.
AgentDojo · verified
Zero false positives, ~1x overhead. Public benchmark, workspace suite, tool_knowledge attack.
attack success, ungoverned to Signet on
0 of 560 in the two latest verified runs · 1 of 560 worst case
| Metric | Ungov. | Signet |
|---|---|---|
| Attack success | 18.39% | <0.2% |
| Utility, under attack | 48.39% | 53.04% |
| Utility, clean tasks | 78.6% | 71.4% |
No legitimate call was blocked by the gate. Every deny was an attack-shaped action.
gpt-4o-2024-05-13. workspace suite, tool_knowledge attack, 560 combinations per pass, three arms. Verified from raw output and the hash-chained ledger across three independent runs. Below Progent's ~1%.
Agent Security Bench · DPI, verified
Real work still got done under ever-increasing attack vectors.
DPI attack success, ungoverned to Signet on
300 of 300 blocked · 3 agents, 5 DPI variants
| DPI variant | Ungov. | Signet |
|---|---|---|
| naive | 58.3% | 0.0% |
| fake_completion | 63.3% | 0.0% |
| escape_characters | 66.7% | 0.0% |
| context_ignoring | 91.7% | 0.0% |
| combined_attack | 91.7% | 0.0% |
As attacks escalate (58% to 91.7% ungoverned), the Signet column stays 0.0% straight down.
gpt-4o-2024-05-13. 3 agents × 5 DPI variants × 20 attacker tools = 300 attacks per arm. Verified from raw run trees and hash-chained ledgers. OPI, memory-poisoning, and aggressive-tool tiers pending.
The field
Signet (TrajeXY), Progent (academia), CaMeL (Google), FIDES (Microsoft).
| Benchmark | Ungoverned | With Signet | Best published |
|---|---|---|---|
| AgentDojo, attack success | 18.39% | <0.2% | 1.0% (Progent) |
| Agent Security Bench, attack success | 74.3% | 0.0% | 3.9% (Progent) |
| Utility retained, under attack | 48.39% | 53.04% | maintained |
| Token / latency overhead | n/a | ~1.0x, flat | up to 2.82x (CaMeL) |
The full diligence pack — raw runs, sealed ledgers, and owned diffs — is available on request.