HipCortex Memory Engine & Cognitive OS for VS Code & Antigravity IDE (`v0.5.8`)

Give your AI coding assistant (VS Code Copilot & Antigravity IDE) persistent, cross-session causal memory, world-model prediction/rollout, and topological graph tools.

Product server / pip / npm are 0.5.2; this extension package is 0.5.8 (10 LM tools, dual /health, multi-OS bundled webserver, chmod +x on bundled Mac/Linux binaries). See docs/channels.md.

Zero-config onboarding (no Rust or Cargo required)

Install from Marketplace / Open VSX / GitHub release VSIX. Extension starts a local Rust webserver under ~/.hipcortex-vscode/bin/ (or uses hipcortex.apiUrl).

Zero external DB / Docker for default petgraph path
Local-first storage under ~/.hipcortex-vscode/storage
Dual /health: accepts plain ok and JSON {status,service} (macOS-friendly ready poll)
Executable bundled bins: ensureUnixExecutable applies chmod 0755 on darwin/linux for bundled, CLI, cache, and download paths (fixes spawn EACCES on fresh Mac install)
Auto-recovery: restarts server before queries when down
Release install: build/package hipcortex-memory-0.5.8.vsix (or latest GitHub Release VSIX)

code --install-extension hipcortex-memory-0.5.8.vsix

`@hipcortex` chat commands

Open Copilot / Antigravity chat and type @hipcortex:

@hipcortex health — server status, tier counts, Merkle integrity
@hipcortex add <content> — store decision / preference / constraint
@hipcortex query <query> — semantic + topological retrieval
@hipcortex status — Headroom vs Caveman mode and savings

Language Model Tools (10)

Extension registers 10 tools with vscode.lm (requires host LM tool API):

Tool	Purpose
`hipcortex_search`	Semantic + live_beliefs-aware search
`hipcortex_health`	Health + capability gate
`hipcortex_predict`	WorldModel single-step `P(s'\|s,a)`
`hipcortex_rollout`	Multi-step Dirichlet / MCTS rollout (`POST /worldmodel/rollout`)
`hipcortex_graph_search`	PPR / related memories from seed UUID
`hipcortex_causal`	Causal attribution
`hipcortex_topo_ppr`	Topo graph Personalized PageRank (`/topo/*`)
`hipcortex_deconstruct`	Hypothesis → candidate edges (`/topo/deconstruct`)
`hipcortex_check_edge`	Contradiction / cycle check before link
`hipcortex_can_execute`	SelfModel gate

MCP hosts use the Python MCP server (18 tools) via hipcortex install — parallel path; see sdk/mcp/README.md.

Headroom & Caveman (token savings)

Headroom (Top-5): ~59–84% token reduction vs full history dump
Caveman (Top-3): ~70–88% in tight debug loops

Configuration (`settings.json`)

{
  "hipcortex.apiUrl": "http://127.0.0.1:3030",
  "hipcortex.apiKey": "",
  "hipcortex.autoStart": true,
  "hipcortex.optimizationMode": "headroom"
}

Local development & packaging

cd vscode-extension
npm install
npm run compile
npm test
npx @vscode/vsce package --allow-missing-repository

Produces hipcortex-memory-0.5.8.vsix (version from package.json).

CI release workflow packages multi-OS VSIX after matrix binary builds.

Channel honesty: docs/channels.md · hipcortex channels
Capability matrix: docs/capabilities.md
Host wizards: docs/hosts/README.md

HipCortex Memory Engine & Cognitive OS

farmountain

HipCortex Memory Engine & Cognitive OS for VS Code & Antigravity IDE (v0.5.8)