AgentForge Ops
Agents:
0/6
alive
Open findings:
—
Last cycle:
never
Cost (24h):
—
!
Bugs
?
Help
⚡ Trigger attack
Public demo (Tailscale Funnel)
▾
1 · Attack Surface (target ← adversarial)
?
v2
⇄
v1
Left: the Week 2 Co-Pilot we're testing, annotated with TM-01..06 attack categories and trust boundaries. Right: AgentForge's adversarial fleet attacking it.
TARGET
Week 2 Clinical Co-Pilot
PRD Stage 2 · TM-01..06
loading target topology…
←
attacks
ADVERSARIAL
AgentForge agents
Orchestrator · Red Team · Judge · Docs · Patch · Watchdog
loading…
2 · Coverage Heatmap
?
Which attack categories have been tested and how well the platform is defending. Greener = more resisted; redder = more successful exploits.
loading coverage…
3 · Live Attack Feed
?
Every attack the platform launches and the Judge's verdict, in real time. Scrolls fastest during active campaigns.
awaiting attack traffic…
4 · Cost Burndown
?
How much each campaign has cost in LLM tokens and dollars. The dashed line is the kill switch — campaigns halt automatically when they cross it.
loading cost…
5 · Severity Over Time
?
Open vulnerabilities by severity, plotted over time. The trend line tells you whether the target is getting more or less defensible.
loading severity…