IDE-style navigation for structured data — code, JSON, YAML. Jump to definitions, find callers, follow references. Available as an MCP server or a mounted folder.

Agentic research control plane: queue state, worker preflight, wake-gated execution, evidence sync, dashboard, alerts, and AI-generated paper packaging.

Open-source data plane primitives: arena storage, schema contracts, filesystem projection, tree-sitter AST, LSP integration

📿 Cross-repo task orchestrator — weaves beads, Linear, and review layers into coordinated work

Claude Code research ladder: 6 composable skills (L0→L5) from 10s fact-check to 1h knowledge vault, plus /osint-research for entity recon. Stack: Firecrawl (search+scrape+summary) + Tavily (search + research with structured schema verdicts) + Exa (neural) + optional Codex CLI cross-model. MIT.

Local-first agentic research platform that turns project bundles into evidence-bound scientific papers with deterministic gates, scientist-panel review, fault localization, and rollback.

Signet: An offline-first identity and authentication protocol designed to replace insecure bearer tokens with phishing-resistant, proof-of-possession credentials

Metaventions AI Research Framework — Multi-tier signal capture for frontier intelligence. Architected Intelligence.

Self-sovereign agent identity. Cap'n Proto schemas, CF Worker authority, ley-line-sign Wasm.

A domain-agnostic autonomous research engine. Feeds in, verified outputs out, with an eval flywheel at the core that turns operating the system into a labeled training dataset.

Agent-agnostic autonomous research protocol adapted from Deli AutoResearch, with STORM-style multi-perspective questioning, source-grounded claim ledgers, persistent state, Discovery Tail Pass, and sufficiency checks for long-horizon AI agents.

🛡️ A modern Go library for CMS/PKCS#7 messages with Ed25519 support.

Go crypto.Signer implementations backed by hardware tokens (PKCS#11 — YubiKeys, HSMs) and macOS Secure Enclave (Touch ID). Build-tag gated, cgo-based; consumed by signet for offline-first signing.

An agentic research project to collect, document, and track progress in longevity science

Standing Eve agent and MCP server for scheduled PR, issue, and snapshot status boards.

Compact LLM lab (MoE, multimodal, FlashAttention) with an agentic autoresearch loop — AI agent iterates on core/ via 5-min experiments on 1×H100. 3 sessions, $13.22, 22 experiments, 5.6% val_bpb improvement. See dev/auto_findings/.

Contract-gated verification for multi-agent research pipelines. Validates every artifact, walks the provenance hash chain, checks the reviewer was not the producer, and returns an exit code. Orchestrates nothing, calls no model API.

Continually improving paper writing skill to transform my experiments into research papers utilising my tone / tone of well written papers I like.

Online goal-reaching RL with diffusion planning: a 2D point-mass prototype plus a Maze2D Diffuser workflow driven by an autonomous agentic experiment controller.

Agentic, visualization-first ML/Math research framework for actuaries & data scientists — a reconnaissance aircraft for research