MCP server: heaptrm as Model Context Protocol tool

Risk scoring in heaptrm tool: pre-corruption exploit prediction

Defensible benchmark: heaptrm 65% vs blind 32% same-session

Benchmark results: observe-once fails (stale addresses across processes)

20-challenge suite benchmark: heaptrm 65% vs blind 25% overall

Benchmark: heaptrm 100% vs blind 0% on hard_heap (no UAF read, heap noise)

Verified: LLM-driven tcache poison with safe-linking bypass via heaptrm

Quarantine off by default, fix false tcache_poison primitives

Add exploit demo: step-by-step tcache poison through heaptrm tool

heaptrm: address extraction, exit detection, cleaner response format

heaptrm: check command, stdout capture, SIGUSR1 validation

heaptrm CLI: compiled Rust binary for LLM-assisted heap exploitation

Add HeapObserver: structured heap state for LLM exploit generation

CVE benchmark: 4/5 real CVE patterns detected, 0 false positives

HeapTRM fuzzer: mutation-based fuzzing guided by heap corruption detection

Tuned verdict logic: F1=1.0 on CVE patterns, zero false positives

Quarantine UAF detection: 5/5 CVE patterns, 0 false positives

Add heaptrm package: v2 harness, CLI, pwntools integration, CVE tests