The smelt MCP server: structure-aware, reversible context optimization as five stdio tools.

Short, obfuscated, collision-proof, reversible identifiers

A reversible GitHub-native multi-agent development pipeline CLI

CCR reversible tool-output compression: compress large grep/log tool results, cache the original, retrieve it via context_retrieve

apeman component with reversible effect.

Sensitive Personal Information (SPI) Protecter — local reversible privacy redaction for Pi

Encode and decode data to string using mustache-like template strings

Zap kernel: composable plugin contexts with reversible effects, services, events, and loader reconciliation.

Reversible PII redaction (pseudonymization) for LLM pipelines. Zero dependencies. Streaming restore.

Reversible, structure-aware context compression for LLM agents — the tokfold command-line tool

A reversible base62 ID obfuscator

baseH: reversible, checksummed, human-safe short references for integers.

One reversible configuration fabric for Codex, Claude Code, and OpenCode.

Keystone 6 Reversible encryption field

Fast deterministic reversible ID encoding for numbers and strings using AES-128-CTR with base62 output

Wire @designless/annotate into a Next.js or Vite-React project so Designless can route a rendered element back to its source line. One command, dev-only, reversible.

One-tool code mode for Pi with parallel context, reasoning checkpoints, and reversible edits

Deterministic, reversible in-session context compaction for the Pi coding agent — folds stale tool output out of the model's view, never out of the session.

Reversible execution for AI agents: whole-workspace checkpoint/rewind, a refuse-and-record effect barrier, and a tamper-evident SHA-256 evidence chain — pure logic over logs and traces.

A global coding-discipline spec for Codex with native hooks, reversible global AGENTS.md installation, and rule-specific opportunity telemetry.