jhammant

MCP server for searching and surfacing Claude Code conversation history

Spin up any open LLM on rented GPUs (vast.ai/RunPod), serve it with vLLM or llama.cpp, and drive it from Claude Code via Claude Code Router. Pay by the hour.

Turbo1Bit: Combining 1-bit LLM weights (Bonsai) with TurboQuant KV cache compression for maximum inference efficiency. 4.2x KV cache compression + 16x weight compression = ~10x total memory reduction.

AI-powered learning platform for Ilya Sutskever's legendary AI paper reading list. Master the foundations of modern AI with curated paths, AI tutoring, and gamification.

Free, open-source 11+ exam practice - 1,364 verified questions for grammar school preparation

A private, on-device macOS rewind: capture, search, and scrub back through your day. 100% local, no telemetry.

🔮 Sub-500ms voice agent with speculative pre-generation. The more you talk, the faster it gets.

Stress-test AI agents for goal drift and system prompt violations. Inspired by 'Asymmetric Goal Drift in Coding Agents' (ICLR 2026).

🦞 Remote coding agent orchestrator — manage Claude Code, Codex, Aider sessions via SSH+tmux

AI agent team setup with OpenClaw - 7 specialized agents working together

Discover your travel history from Apple Photos and get AI-powered holiday recommendations

No description.

What 400+ world-class performers agree on — AI-powered analysis of every Diary of a CEO episode. Interactive knowledge graph, cross-episode patterns, and key learnings.

No description.

Graph-based self-healing tool routing for LLM agents. 93% fewer LLM calls than ReAct.

No description.

Loom — agent-native, additive app host. Deploy small apps: instantly routed, discoverable, agent-callable (MCP), service-consuming. BSL 1.1 open core.

No description.

MCP server that gives Claude Code native Discord tools — bridge to remote AI agents

⚡ Dynamic MoE expert streaming for Apple Silicon — run 685B parameter models on a laptop. Rust-native GPU + ANE + NVMe orchestration.

Claude Code quota intelligence: live limits, burn-rate forecasts, API-cost equivalence, and capacity advice for agents

AI agent dreaming system — replay, consolidation, creative synthesis during quiet hours

No description.

A Claude Code skill that delegates self-contained coding tasks to Codex (gpt-5.6-sol) — background, sandboxed, on a separate quota pool, with diff review.

See the algorithm that's feeding you — a free, open tool that reads your social feed and shows what it's reinforcing. doomscrollstats.com

LLM-assisted security scanner: network + code, AI triage, continuous pen-testing (free & OSS)

Analyze your AWS Bedrock spend and recommend cheaper providers — before you route a single call. Analyze -> replay-eval -> govern.

Intelligent clipboard for Claude Code: /clp smart-copies your last deliverable, /pst reads your clipboard and acts on it in context. Cross-platform.

Deterministic codebase analysis (MCP + Taxi) as ground truth for LLM agents — every signal carries file:line evidence.

LLM forecasting toolkit: Halawi retrieval + multi-agent simulation + Kelly accumulator. Geopolitics, box office, sports. 100% local on M5 Mac.

Local-LLM-powered security scanner. Multi-lens per-file review + architecture extraction + threat-model synthesis + sandboxed exploit PoCs. 100% local inference via LM Studio.

Benchmark and training data generation for LLM-based Infrastructure-as-Code repair via environment inversion

Modified USAF Aircraft DATCOM

👻 Ghostty is a fast, feature-rich, and cross-platform terminal emulator that uses platform-native UI and GPU acceleration.

Profile and optimize hot loops on ARM (Neoverse-N1 / Apple Silicon). Finds decode bottlenecks, cache misses, branch mispredictions.

Profile and optimize hot loops on x86 (Intel DSB + AMD OpCache). Find decode bottlenecks, uOP cache misses, branch mispredictions.

No description.

No description.

Provenance + license checker for model weights — hash-based identity that fails CI on unknown, unpinned, or non-compliant model artifacts

Scan a series of commits/PRs for a malicious capability assembled across many small diffs — distributed-attack detection for AI coding agents (arXiv 2607.02514)

Compile fuzzy natural-language functions into tiny local models — zero per-call API cost at runtime (arXiv 2607.02512)

Map the cryptography across your live estate (TLS, SSH, X.509) and rank it against a quantum threat model. Emits a CycloneDX CBOM + SARIF. Stdlib-only Go.

Does messy spelling break AI? Open benchmark of how well LLMs understand dyslexia-style misspelled input, with an interactive predictor. Neurodiversity edition.

A benchmark for how LLMs handle benign political satire: over-refusal vs safety calibration, across text and image models.

Pit LLMs against each other at heads-up No-Limit Hold'em — Elo + BB/100 ratings, a strategy-docs 'tuition' experiment, and a live poker broadcast.

Tap It — website, privacy policy & support

Discover public APIs and map them to Taxi taxonomies for OrbitalHQ

GitHub user/repo fingerprints for humans, plus agent-ready repo context packs

Template for customising your devprint.dev card. Fork or use-as-template, edit devprint.json, push.

Generate A PDF report from a watchlist of UK equities. The report displays Price/Volume, Stochastic and MACD charts as well as some selected fundamental data.

Bootstrap any AI agent (Claude Code, Cursor, Codex…) from one URL. boothub.dev/USERNAME → manifest → swarm.

jhammant's boothub reference profile. Bootstraps via boothub.dev/jhammant — skills, MCPs, swarm, all the things.

Interactive explainer on AlphaFold and proteome-scale structure prediction

Interactive explainer on AlphaFold and AI for biology

Interactive explainer: Folding Time, AlphaFold, and proteome-scale structure prediction

Interactive explainer for Superposition Yields Robust Neural Scaling

No description.

Fine-tuning Gemma 3 4B into a UK 11+ tutor — three LoRAs, one surprise winner

Interactive guide to Assembly Theory's physics of causation

An interactive guide to Assembly Theory: count the steps, see the fight.

An interactive relativity explainer — 4.37 light-years to Alpha Centauri in 10 minutes flat

Interactive microsite: how open-weight LLMs improved dramatically on the same consumer Mac hardware from 2023-2026, plus projections to 2028

No description.

Interactive quantum computing explorer for World Quantum Day 2026. Bloch sphere, quantum gates, Grover's, Shor's, optimization, and QKD visualizations.

Analyze blink rates in videos to detect psychopathic traits. Search YouTube by name, cross-video face matching, CNN eye classifier, psychopathy risk assessment.

Advanced analytics platform for video content analysis with interactive dashboards and network visualizations

Open-source hybrid fact verification using Knowledge Graphs (Wikidata) + Web Search. Inspired by WKGFC (arxiv 2603.00267).

Educational video demonstrating TurboQuant vector quantization algorithm with Python implementation

AI-powered analysis of every Acquired podcast episode — business, tech, and M&A insights

AI-powered analysis of every Dwarkesh Podcast episode

AI-powered analysis of 2,600+ Joe Rogan Experience episodes — insights, consensus, contrarian takes

AI-powered analysis of every TED Talk — insights, advice, and contrarian views from 4,600+ speakers

UK political party donations visualiser — Electoral Commission data 2022-2026

No description.

An LLM generated grad level deep learning text book

Interactive timeline showing zackpolanski.com's evolution from hypnotherapy to Green Party leader (2009-2026)

Interactive visualization of the first confirmed runaway supermassive black hole (Van Dokkum et al. 2026)

Deep Learning and Transformers - A Graduate-Level Textbook

Interactive D3.js visualization of Space Coast rocket facility scale — SpaceX, Blue Origin, Amazon Kuiper, AST SpaceMobile

Where Do MPs Get Their Money? Interactive visualization of UK Parliament financial interests

Find and disable hallucination neurons in LLMs. Reproduction of arxiv.org/abs/2512.01797

Satellite image analysis with Meta SAM2 — auto-segment buildings, roads, ships, water from overhead imagery

Benchmark suite comparing biological neural networks (Cortical Labs DishBrain) against conventional AI baselines

go-audit is an alternative to the auditd daemon that ships with many distros

Gym Scraper for Better Gyms

OpenShift 3 - build, deploy, and manage your applications with Docker and Kubernetes