Researchers

Research tooling

Lab tools for empirical study. Start from the researchers hub for the full audit path. Findings: B3 report.

Experimental cryptography only. Green tests and clean mass campaigns show consistency with the v1 specification — not production safety.

Research report

Living notes: construction summary, conformance, collision campaign log (100 000 random pairs, 0 collisions), TMTO snapshot, risks, open questions.

Open research report docs/RESEARCH_REPORT.md

Collision Lab

Windows WPF app for mass uniqueness hunts (multi-worker).

dotnet run --project src/ForgeHash.CollisionLab -c Release

Kinds: distinct passwords/salts, random pairs, bit-flips, parameter sets, truncated outputs. Export JSON/CSV. Prefer Development (8 MiB) for large N on B3. For ForgeHash-X, choose Algorithm ForgeHash-X and the Toy preset (m=1024) — throughput is not comparable to B3 Development.

ForgeHash-X research

Separate sandbox with a custom ForgeX sponge. Notes, paper PDF, matched-cost throughput benches, and toy-vector campaigns live under docs/forgehx/.

RESEARCH_NOTES.md X research paper (PDF) X Vectors SPECIFICATION_X.md ResearchBench (B3 vs X)

dotnet run --project tools/ForgeHash.ResearchBench -c Release -- --suite peers --markdown
dotnet run --project tools/ForgeHash.ResearchBench -c Release -- --suite matched --markdown

Peer suite times bcrypt, Argon2id, scrypt, and PBKDF2 at documented presets next to ForgeHash-B3/X. Presets are not equal-work; see RESEARCH_NOTES §3b.

Visualizer

dotnet run --project src/ForgeHash.Visualizer -c Release -- all --out artifacts/analysis --memory 8192 --iterations 1 --parallelism 2

Writes report.json, reference CSV, heatmap, lanes, Graphviz DOT, TMTO JSON, and a short notes file.

Scope

Topic	Status
Official v1 vectors	Frozen
Collision campaigns	CI + GUI (1e5 random pairs, 0 hits logged)
Avalanche / memory / references	Automated
TMTO stride heuristic	Diagnostic only
GPU / ASIC / side-channel lab	Open

docs/RESEARCH.md ForgeHash.Analysis