mode Mythos

mythos-cyber-bench-survey-mode

Survey of cyber-eval benchmarks (NYU CTF, InterCode-CTF, Cybench, CyberSecEval, AISI ranges) with saturation curves and a "what to use when" decision matrix

View source on GitHub ↗ ← Back to search
KindMode
CategoryMythos
Installnpx -y github:anubhavg-icpl/vibe add mythos-cyber-bench-survey-mode
LicenseCC BY-NC-SA 4.0

More in Mythos

mode

mythos-behavioral-analysis-mode

Automated transcript analysis of autonomous cyber-agent runs — unique services, exploit/exploration ratio, credential reuse, drift detection, cost-per-milestone

View → mode

mythos-context-compaction-eval-mode

Implement and audit context compaction for long-horizon agent runs — ~80% trigger, summarization fidelity, KV-cache cost tradeoffs, credential handling

View → mode

mythos-cyber-eval-disclosure-mode

Responsible disclosure norms for cyber-eval results — methodology without exploit recipes, hash-then-reveal pattern, coordinating with developers and governments

View → mode

mythos-cyber-range-designer-mode

Design multi-step, milestone-graded cyber ranges for AI agent evaluation, modeled on AISI's "The Last Ones" and "Cooling Tower"

View → mode

mythos-inference-scaling-eval-mode

Run token-budget sweeps against autonomous cyber agents to measure log-linear inference-time compute scaling, modeled on AISI's 10M→100M findings

View → mode

mythos-inspect-ai-harness-mode

Build agent evaluations on the UK AISI Inspect AI framework with Sandboxing Toolkit integration, real Task/Solver/Scorer APIs

View →