Start a fresh pi session seeded with this conversation's flattened transcript. No LLM summary, no tokens burned, nothing paraphrased.

No description provided.

MCP server giving AI agents (Claude Code, Claude Desktop, Cursor, ChatGPT, Codex, OpenClaw) persistent long-term memory backed by your local Obsidian markdown vault. Hybrid retrieval (BM25 + ML embeddings + BGE reranker, RRF-fused), HNSW + int8 quantizati

Agent security runtime for Node.js and browsers: provenance-tracked tool-call guard (CaMeL-style), spotlighting, fuzzy canaries, and hybrid rules + embedded ML detection for prompt injection. Zero dependencies, TypeScript-first, published benchmarks.

Local transformers.js embedding provider for @vivantel/rag-core

An MCP server for an open-licensed corpus, served with verifiable provenance — every document carries its sha256, DOI and OpenTimestamps proof so a consuming agent can check its own citation.

One memory for every AI coding agent — plain markdown in your git. Decisions, projects, ledger & board together; recall@5 0.925; 12 engines via install.

Sense — universal text embedding library for TensorFlow.js (USE lite + multilingual transformers)

Provenance labeling for agent inputs, enforced at the tool-call boundary. Blocks untrusted data from reaching shell, filesystem, and network sinks. Ships with an injection corpus and a published list of known gaps.

<picture> <source media="(prefers-color-scheme: dark)" srcset="assets/logo-negro.jpeg"> <source media="(prefers-color-scheme: light)" srcset="assets/logo-blanco.jpeg"> <img alt="ragnarok logo" src="assets/logo-blanco.jpeg" width="100%"> </picture>

VCC-style instant, near-lossless deterministic compaction engine for the DeepSeek Harness — a drop-in replacement for @deepseek-ai/dsh-compaction-basic

Production caching seam for AI workloads (keyless/offline memory reference + Redis adapter): exact-match cache, verified semantic cache, embedding cache, singleflight request coalescing, negative cache, and epoch-based invalidation. Makes the gateway + RA

Narrative state and character knowledge for AI game masters, with entity resolution, atomic turn commits, and a JSON CLI.

Canonical book-digitizer skill pack — 8 skills that turn an engineering book/textbook/monograph into an installable skill pack whose skills make AI coders work within the book's methodologies: digitize-book (orchestrator), book-ingest, book-knowledge-extr

tokenolo — authenticated CLI for the tokenolo token-optimization API. Rewrites prompts to be token-cheaper while preserving intent and every constraint.

Deterministic-first, plain-markdown knowledge vault. Code does the bulk transform. The LLM only touches the irreducibly-semantic 20%. Native grep plus BM25 retrieval, zero MCP on the vault.

Portable skill and CLI for removing AI sounding fluff from decks, docs, PDFs, and notes while keeping the facts.

Names PR-review sessions after the pull request's real subject, so the session list and terminal tab read "PR review 1234: <PR title>".

Safe-C coding plugin for Hydron — MISRA C:2012-aligned defensive C/C++ guidelines as an always-on skill plus a safe-c-reviewer subagent

CueMap TypeScript SDK - High-performance temporal-associative memory