A CLI tool to build context files to attach to Large Language Models (LLMs).

Compress LLM context. Save tokens. Fit more into every request.

Liminal — transparent LLM context compression proxy

Optimize token allocation within LLM context windows

Lossless, provable, composable LLM context compression

Extract video frames for LLM context injection

Pre-computed LLM context for AI coding agents.

Universal context intelligence layer — compresses LLM context across CLI, MCP, browser, and IDE

Persistent structured memory for AI agent systems — pluggable SQLite or Postgres backend, LLM context bridge

Recursive Language Model - Process documents larger than LLM context windows

Pi extension for bounded Brave web search and LLM context retrieval

Emit an llms.txt-style, token-efficient digest of the cssdoc model for LLM context windows.

In-process content + symbol cache for LLM context windows. Keyed (source, version, item). Tree-sitter symbol extraction + grep + cached-read bridge as subpath imports.

Read-only MCP server and LLM context generator for Deweyou Design.

Rehydra plugin for OpenCode — anonymizes secrets in LLM context using native plugin hooks

Pi extension: inspect the system prompt and full LLM context in scrollable overlay popups

Transparent LLM context compression for Pi using Headroom

A DSL for explicit, scoped, auditable LLM context

Kern Project Context Graph — compact, deterministic whole-project map for LLM context

SuperCompress — one install, all your coding agents. Auto-compresses LLM context to save ~65% on token costs.