@hybridlabor-api/aos 4.11.0 → 4.12.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.de.md +296 -725
- package/README.md +294 -831
- package/README.pt.md +316 -677
- package/bin/aos-store.mjs +29 -59
- package/installer.js +39 -2
- package/lib/ecc-store-index.json +1 -1
- package/lib/store-shared.mjs +133 -0
- package/lib/store-ui/index.html +150 -0
- package/lib/store-ui/server.mjs +180 -0
- package/package.json +2 -2
- package/scripts/build-ecc-store-index.mjs +17 -1
- package/scripts/build-plugin-manifest.mjs +3 -3
- package/skills/bdb-aos/SKILL.md +46 -0
- package/skills/global_config/aos-store/SKILL.md +71 -0
package/README.md
CHANGED
|
@@ -1,708 +1,347 @@
|
|
|
1
1
|
|
|
2
2
|
|
|
3
3
|
🌐 **Language / Sprache / Idioma**: **English** | [ 🇩🇪 Deutsch ](README.de.md) | [ 🇵🇹 Português ](README.pt.md)
|
|
4
|
-
# 🚀 AOS — BDB Agent OS · Optimized Creative & Full-Stack Skills Pack
|
|
5
4
|
|
|
6
|
-
|
|
5
|
+
# AOS — BDB Agent OS
|
|
6
|
+
|
|
7
7
|
[](https://www.npmjs.com/package/@hybridlabor-api/aos)
|
|
8
8
|
[](https://www.npmjs.com/package/@hybridlabor-api/aos)
|
|
9
|
-
[](LICENSE)
|
|
10
9
|
[](https://github.com/hybridlabor-api/aos/stargazers)
|
|
11
10
|
[](https://github.com/hybridlabor-api/aos/commits/main)
|
|
12
|
-
|
|
13
|
-
[](https://github.com/hybridlabor-api/aos/actions)
|
|
12
|
+
[](LICENSE)
|
|
13
|
+
[](package.json)
|
|
14
|
+
[](#skills)
|
|
15
|
+
[](#mcp-servers)
|
|
16
|
+
[](#supported-harnesses)
|
|
17
17
|
[](https://github.com/NVIDIA/SkillSpector)
|
|
18
18
|
|
|
19
|
-
|
|
19
|
+
AOS installs a curated skill library, a subagent roster, gate hooks and a runnable multi-agent build pipeline into every coding-agent harness on your machine.
|
|
20
20
|
|
|
21
|
-
|
|
21
|
+
```bash
|
|
22
|
+
npx -y @hybridlabor-api/aos@latest
|
|
23
|
+
```
|
|
22
24
|
|
|
23
|
-
**
|
|
25
|
+
Built for people who already run **Claude Code, Google Antigravity, Codex CLI, OpenCode, Cursor, Windsurf, Roo Code / Cline or Aider** and want all of them to behave the same way.
|
|
24
26
|
|
|
25
|
-
|
|
27
|
+
After install you have:
|
|
26
28
|
|
|
27
|
-
|
|
28
|
-
|
|
29
|
+
- **<!-- count:skills -->213<!-- /count --> skills** in six categories, discoverable by every harness as `<name>/SKILL.md`.
|
|
30
|
+
- **<!-- count:agents -->13<!-- /count --> subagents** (Architect, TechLead, Reviewer, the Godmodes, security and silent-failure reviewers) compiled into each harness's native agent format.
|
|
31
|
+
- **<!-- count:mcps -->21<!-- /count --> MCP servers** for creative software, OS control, memory and cross-harness delegation.
|
|
32
|
+
- **Three pipelines** — `/startcycle`, `/startcycle-graph`, `/startcycle-graph-user` — and a **GO gate** that mechanically blocks `git push`, `npm publish`, `npm version` and recursive `rm`.
|
|
33
|
+
- **Tools:** Plan Canvas, agenttrail, archify, the AOS Store, the Launchpad dashboard and `aos doctor`.
|
|
29
34
|
|
|
30
35
|
---
|
|
31
36
|
|
|
32
|
-
|
|
33
|
-
|
|
34
|
-
This release moves the multi-agent pipeline from prose into an executable
|
|
35
|
-
state machine, and hardens the installer around it.
|
|
36
|
-
|
|
37
|
-
- **A dispatcher graph that actually runs.** Seven nodes, explicit edge
|
|
38
|
-
predicates, a Reviewer repair loop with a no-progress guard, and automated
|
|
39
|
-
escalation to a human when the loop stops making progress. [Details below](#-aos-the-dispatcher-graph).
|
|
40
|
-
- **Three pipeline variants** (`/startcycle`, `/startcycle-graph`,
|
|
41
|
-
`/startcycle-graph-user`) so the machinery matches the task instead of
|
|
42
|
-
forcing full ceremony on a two-file change.
|
|
43
|
-
- **A mechanically enforced GO gate.** `git push`, `npm publish`,
|
|
44
|
-
`npm version` and recursive `rm` are blocked by a `PreToolUse` hook unless
|
|
45
|
-
your immediately preceding message is the literal word **GO** — enforcement
|
|
46
|
-
that survives permission-mode changes, because it is a hook rather than a
|
|
47
|
-
rule an agent is asked to respect.
|
|
48
|
-
- **Verified daemon startup.** The installer no longer reports "service
|
|
49
|
-
started" on faith; it connects to the port and warns you if the daemon
|
|
50
|
-
never came up. Same for the ecosystem status table, which now compares
|
|
51
|
-
versions with real semver precedence across every dist-tag instead of
|
|
52
|
-
string inequality against `latest`.
|
|
53
|
-
- **Separated MCP stores per harness.** Claude Desktop and Claude Code read
|
|
54
|
-
different files; installing for one no longer silently skips the other.
|
|
55
|
-
Existing servers in either file are merged, not overwritten.
|
|
56
|
-
- **Scriptable, non-interactive installs.** `--platforms=<n[,n]>` selects
|
|
57
|
-
targets without the menu, so a machine that only runs Claude Code can be
|
|
58
|
-
provisioned in CI without inheriting an Antigravity-first default.
|
|
59
|
-
|
|
60
|
-
### 🪐 Universal Agent Harness
|
|
61
|
-
The installer features a fully automated Universal Sync engine. It scans your system for **Claude Desktop, Cursor, Windsurf, Aider, Roo/Cline**, and injects the curated MCP configuration and Godmode rules across all environments simultaneously.
|
|
62
|
-
- **Local Project Harness:** Instead of installing globally into `$HOME`, drop the `.agents` contract, the gate hooks and the dispatcher workflow directly into a single project — `npx @hybridlabor-api/aos --project-harness`.
|
|
63
|
-
|
|
64
|
-
### 🧩 Ecosystem Integrations
|
|
65
|
-
This package acts as the bridge to four major upstream capabilities:
|
|
66
|
-
- **BDB AO — Agent Orchestrator:** The orchestration layer for parallel AI agents. Start multiple isolated agent sessions via Git worktrees with live terminal control, automatic CI/CD feedback loops, and PR review routing. Ships one prebuilt binary today (macOS, Apple Silicon); other platforms build from source. Supersedes the now-archived `bdb-os-agent-workspace` repository — do not install that one.
|
|
67
|
-
- **BDB Creator Extension:** The heavy-lifting Agentic Media Pipeline. Gives agents local ComfyUI MCP capabilities (FLUX, SDXL), Image-to-3D generation (TripoSR, TRELLIS), and automated video production through OpenMontage and Remotion.
|
|
68
|
-
- **BDB Synapse:** 3D Codebase Visualization & Agent Session Replay. Renders your repository as an interactive code city and replays agent sessions as light trails, showing which files were read, edited, and where friction occurred.
|
|
69
|
-
- **BDB Hardware & PCB:** Optional electrical/PCB design module. Two local MCP servers (KiCad, OpenSCAD) exposing 55 tools, plus 5 skills covering schematic capture, layout/routing, and DFM sign-off. Install standalone or via this installer's optional-module picker; paired with the core `godmode-hardware-pcb` persona.
|
|
70
|
-
|
|
71
|
-
### 🔗 Recommended Companion Plugins (Claude Code)
|
|
72
|
-
Neither of these ships inside this package — they are independent, community-maintained
|
|
73
|
-
Claude Code plugins that pair naturally with the `agy`-delegation pattern
|
|
74
|
-
[`/startcycle-graph-user`](#-aos-the-dispatcher-graph) already uses. Install them
|
|
75
|
-
separately if you want the same routing available outside a `/startcycle` run.
|
|
76
|
-
|
|
77
|
-
- **[antigravity-for-claude-code](https://github.com/yuting0624/antigravity-for-claude-code)**
|
|
78
|
-
— runs the Antigravity CLI (`agy`, Gemini) as a collaborating sub-agent with
|
|
79
|
-
intelligent model routing across the SDLC.
|
|
80
|
-
```bash
|
|
81
|
-
claude plugin marketplace add yuting0624/antigravity-for-claude-code
|
|
82
|
-
claude plugin install antigravity@antigravity-for-claude-code
|
|
83
|
-
```
|
|
84
|
-
- **[opencode-plugin-cc](https://github.com/tasict/opencode-plugin-cc)** — adds
|
|
85
|
-
`/opencode:review` / `/opencode:adversarial-review` slash commands, letting Claude Code
|
|
86
|
-
delegate an async job to OpenCode and set up a review gate that blocks progress until
|
|
87
|
-
OpenCode's review comes back clean.
|
|
88
|
-
```bash
|
|
89
|
-
claude plugin marketplace add https://github.com/tasict/opencode-plugin-cc.git
|
|
90
|
-
claude plugin install opencode@tasict-opencode-plugin-cc
|
|
91
|
-
```
|
|
92
|
-
|
|
93
|
-
> [!CAUTION]
|
|
94
|
-
> If you also have an older `antigravity-plugin-cc` marketplace installed, disable it and
|
|
95
|
-
> clear its plugin cache first — two `agy`-routing plugins active at once causes a
|
|
96
|
-
> dual-activation conflict where neither initializes cleanly.
|
|
97
|
-
|
|
98
|
-
## Overview
|
|
99
|
-
|
|
100
|
-
This repo ships three things: a curated skill library for coding agents, an
|
|
101
|
-
installer that wires them (plus 19 local MCP wrappers) into whichever harness
|
|
102
|
-
you use, and a dispatcher graph that orchestrates them as a multi-agent
|
|
103
|
-
build pipeline. See [AOS: The Dispatcher Graph](#-aos-the-dispatcher-graph)
|
|
104
|
-
below for how the pipeline itself works.
|
|
37
|
+
## The Dispatcher Graph
|
|
105
38
|
|
|
106
|
-
|
|
39
|
+
The graph is harness-neutral and runs on Claude Code's Dynamic Workflows, Antigravity's parallel execution, and any compatible agent harness.
|
|
107
40
|
|
|
108
|
-
|
|
41
|
+
```mermaid
|
|
42
|
+
flowchart LR
|
|
43
|
+
U(["👤 User"])
|
|
44
|
+
A["<b>Architect</b><br/><span>System Plan</span>"]
|
|
45
|
+
T["<b>TechLead</b><br/><span>Capability Map</span>"]
|
|
46
|
+
UX["<b>UI_UX</b><br/><span>Frontend</span>"]
|
|
47
|
+
EN["<b>Engineering</b><br/><span>Backend</span>"]
|
|
48
|
+
ME["<b>Media_EventTech</b><br/><span>Creative</span>"]
|
|
49
|
+
R["<b>Reviewer</b><br/><span>QA</span>"]
|
|
50
|
+
S["<b>Shipping</b><br/><span>Gate</span>"]
|
|
109
51
|
|
|
110
|
-
|
|
52
|
+
U --> A --> T
|
|
53
|
+
T --> UX & EN & ME
|
|
54
|
+
UX & EN & ME --> R
|
|
55
|
+
R --> S
|
|
111
56
|
|
|
112
|
-
|
|
57
|
+
T -.->|reject| A
|
|
58
|
+
R -.->|findings| UX
|
|
59
|
+
S -.->|gate fail| EN
|
|
60
|
+
R -.->|escalate| U
|
|
61
|
+
```
|
|
113
62
|
|
|
114
63
|
---
|
|
115
64
|
|
|
116
|
-
|
|
65
|
+
## Install
|
|
66
|
+
|
|
67
|
+
Requirements: Node.js >= 20. macOS, Linux and Windows (PowerShell).
|
|
68
|
+
|
|
69
|
+
```bash
|
|
70
|
+
npx -y @hybridlabor-api/aos@latest
|
|
71
|
+
```
|
|
117
72
|
|
|
118
|
-
|
|
73
|
+
**First run.** The installer detects which harnesses are present, asks which to target and which tier (Pro MEDIA or Basic), copies the skills into each harness's skill directory, compiles the subagents, wires the hooks, merges the MCP configuration into each harness's own config file (existing entries are kept), and offers the optional modules listed below.
|
|
119
74
|
|
|
120
|
-
|
|
121
|
-
|---------|---------|
|
|
122
|
-
| **`godmode-engineering`** | Forces Domain-Driven Design, strict TypeScript checks, Clean Architecture, and systematic 5-step debugging triage. Dispatcher's `Engineering` node. |
|
|
123
|
-
| **`godmode-ui-ux`** | The frontend Gold-Standard. Enforces Brand Discovery, Anti-Slop principles, DTCG design tokens, fluid motion physics, and enterprise accessibility. Dispatcher's `UI_UX` node. |
|
|
124
|
-
| **`godmode-shipping`** | The final gatekeeper for production releases. Enforces Spec-Driven Development, pre-launch checks, feature flags, and safe rollbacks. Dispatcher's `Shipping` node. |
|
|
125
|
-
| **`godmode-eventtech`** | Architectural authority for real-time performance, signal flow, protocol routing, and hardware constraints in live show and event technology environments. |
|
|
126
|
-
| **`godmode-3d-creation`** | MCP-First master orchestration for 3D generation, mesh reconstruction, and parametric CAD engineering. Interfaces with local 3D engines and MCP tools. |
|
|
127
|
-
| **`godmode-media-creation`** | MCP-First master orchestration for all media creation pipelines (Video, Timeline Assembly, Beat Sync, Motion Design). Directly interfaces with local media engines and MCP tools. |
|
|
128
|
-
| **`godmode-hardware-pcb`** | Architectural authority for electrical schematics, PCB layout, and OpenSCAD enclosure co-design. Enforces IPC-standard trace/impedance math and a headless KiCad ERC/DRC/DFM gate before anything ships to fabrication. Ships in core; the KiCad/OpenSCAD skills and MCP servers it drives live in the optional [**BDB Hardware & PCB**](#-bdb-hardware--pcb-electrical--enclosure-design-module) module. |
|
|
75
|
+
**Every later run** opens a menu instead:
|
|
129
76
|
|
|
130
|
-
|
|
131
|
-
|
|
132
|
-
|
|
133
|
-
|
|
134
|
-
|
|
135
|
-
|
|
77
|
+
| Menu item | What it does |
|
|
78
|
+
|---|---|
|
|
79
|
+
| Quick Update | Refreshes skills, templates, hooks and installed modules to the version you just ran |
|
|
80
|
+
| Run System Checkup / Doctor | Runs `aos doctor`: dependencies, file placement, daemons, hooks |
|
|
81
|
+
| Drop Local Project Harness | Copies the dispatcher contract into the current directory (see below) |
|
|
82
|
+
| Reconfigure System | Change targets, tier or options |
|
|
83
|
+
| Uninstall AOS | Removes what the installer placed; your data stays |
|
|
136
84
|
|
|
137
|
-
|
|
85
|
+
### Non-interactive
|
|
138
86
|
|
|
139
|
-
|
|
140
|
-
|
|
141
|
-
|
|
142
|
-
|
|
143
|
-
|
|
144
|
-
<summary><strong>🎨 Frontend & UI/UX Design</strong> — 20 skills</summary>
|
|
145
|
-
|
|
146
|
-
| Skill | Description |
|
|
147
|
-
|-------|-------------|
|
|
148
|
-
| `design-spells` | Curated micro-interactions and design details that add "magic" and personality to websites and apps. |
|
|
149
|
-
| `design-taste-frontend` | Use when building high-agency frontend interfaces with strict design taste, calibrated color, responsive layout, and motion rules. |
|
|
150
|
-
| `frontend-design` | Frontend designer-engineer skill — not a layout generator. |
|
|
151
|
-
| `frontend-dev-guidelines` | Senior frontend engineer operating under strict architectural and performance standards. |
|
|
152
|
-
| `landing-page-generator` | Generates high-converting Next.js/React landing pages with Tailwind CSS. Uses PAS, AIDA, and BAB frameworks. |
|
|
153
|
-
| `senior-frontend` | Frontend development for React, Next.js, TypeScript, and Tailwind CSS. Bundle analysis, accessibility, and code quality. |
|
|
154
|
-
| `shadcn` | Manages shadcn/ui components and projects with documentation, usage patterns, and modern design systems. |
|
|
155
|
-
| `spline-3d-integration` | Interactive 3D scenes from Spline.design in web projects, including React embedding and runtime control API. |
|
|
156
|
-
| `tailwind-patterns` | Tailwind CSS v4 principles. CSS-first configuration, container queries, modern patterns, design token architecture. |
|
|
157
|
-
| `ui-component` | Generate UI components following StyleSeed Toss conventions for structure, tokens, accessibility, and ergonomics. |
|
|
158
|
-
| `ui-page` | Scaffold mobile-first pages using StyleSeed Toss layout patterns, section rhythm, and shell components. |
|
|
159
|
-
| `ui-pattern` | Reusable UI patterns: card sections, grids, lists, forms, and chart wrappers using StyleSeed Toss primitives. |
|
|
160
|
-
| `ui-review` | Review UI code for design-system compliance, accessibility, mobile ergonomics, and spacing discipline. |
|
|
161
|
-
| `ui-tokens` | List, add, and update StyleSeed design tokens while keeping JSON sources, CSS variables, and dark-mode values in sync. |
|
|
162
|
-
| `ui-ux-pro-max` | Comprehensive design guide for web and mobile applications. Color palettes, typography, and UX review. |
|
|
163
|
-
| `ux-audit` | Audit screens against Nielsen's heuristics and mobile UX best practices. |
|
|
164
|
-
| `ux-feedback` | Add loading, empty, error, and success feedback states with practical mobile-first rules. |
|
|
165
|
-
| `ux-flow` | Design user flows and screen structure using progressive disclosure, hub-and-spoke navigation, and information pyramids. |
|
|
166
|
-
| `ux-persuasion-engineer` | Behavioral UX specialist applying choice architecture, friction audits, and commitment design. |
|
|
167
|
-
| `wcag-audit-patterns` | Auditing web content against WCAG 2.2 guidelines with actionable remediation strategies. |
|
|
168
|
-
|
|
169
|
-
</details>
|
|
170
|
-
|
|
171
|
-
<details>
|
|
172
|
-
<summary><strong>⚛️ React & Next.js</strong> — 10 skills</summary>
|
|
173
|
-
|
|
174
|
-
| Skill | Description |
|
|
175
|
-
|-------|-------------|
|
|
176
|
-
| `nextjs-app-router-patterns` | Comprehensive patterns for Next.js 14+ App Router architecture, Server Components, and modern full-stack React. |
|
|
177
|
-
| `nextjs-best-practices` | Next.js App Router principles. Server Components, data fetching, routing patterns. |
|
|
178
|
-
| `react-best-practices` | Performance optimization guide for React and Next.js applications by Vercel. |
|
|
179
|
-
| `react-component-performance` | Diagnose slow React components and suggest targeted performance fixes. |
|
|
180
|
-
| `react-patterns` | Modern React patterns and principles. Hooks, composition, performance, TypeScript best practices. |
|
|
181
|
-
| `tanstack-query-expert` | TanStack Query (React Query) — asynchronous state management, mutations, optimistic updates, and SSR integration. |
|
|
182
|
-
| `vercel-ai-sdk-expert` | Vercel AI SDK: Core API, UI hooks (useChat, useCompletion), tool calling, and streaming UI components. |
|
|
183
|
-
| `vercel-deployment` | Expert knowledge for deploying to Vercel with Next.js. |
|
|
184
|
-
| `zustand-store-ts` | Create Zustand stores following established patterns with proper TypeScript types and middleware. |
|
|
185
|
-
| `web-artifacts-builder` | Build powerful frontend claude.ai artifacts. |
|
|
186
|
-
|
|
187
|
-
</details>
|
|
188
|
-
|
|
189
|
-
<details>
|
|
190
|
-
<summary><strong>🗄️ Backend, Databases & APIs</strong> — 14 skills</summary>
|
|
191
|
-
|
|
192
|
-
| Skill | Description |
|
|
193
|
-
|-------|-------------|
|
|
194
|
-
| `api-design-principles` | Master REST and GraphQL API design to build intuitive, scalable, and maintainable APIs. |
|
|
195
|
-
| `api-patterns` | API design decision-making. REST vs GraphQL vs tRPC selection, response formats, versioning, pagination. |
|
|
196
|
-
| `database-design` | Database design principles. Schema design, indexing strategy, ORM selection, serverless databases. |
|
|
197
|
-
| `drizzle-orm-expert` | Drizzle ORM for TypeScript — schema design, relational queries, migrations, and serverless database integration. |
|
|
198
|
-
| `golang-pro` | Master Go 1.22+ with modern patterns, advanced concurrency, performance optimization, and production-ready microservices. |
|
|
199
|
-
| `go-concurrency-patterns` | Go concurrency with goroutines, channels, sync primitives, and context. Worker pools and race condition debugging. |
|
|
200
|
-
| `go-playwright` | Robust, stealthy, and efficient browser automation using Playwright Go. |
|
|
201
|
-
| `microservices-patterns` | Microservices architecture patterns: service boundaries, inter-service communication, data management, and resilience. |
|
|
202
|
-
| `neon-postgres` | Expert patterns for Neon serverless Postgres, branching, connection pooling, and Prisma/Drizzle integration. |
|
|
203
|
-
| `openapi-spec-generation` | Generate and maintain OpenAPI 3.1 specifications from code, design-first specs, and validation patterns. |
|
|
204
|
-
| `postgres-best-practices` | Postgres performance optimization and best practices from Supabase. |
|
|
205
|
-
| `postgresql` | Design PostgreSQL-specific schemas. Data types, indexing, constraints, performance patterns, and advanced features. |
|
|
206
|
-
| `prisma-expert` | Prisma ORM — schema design, migrations, query optimization, relations modeling, and database operations. |
|
|
207
|
-
| `using-neon` | Neon serverless Postgres: autoscaling, branching, instant restore, and scale-to-zero. |
|
|
208
|
-
|
|
209
|
-
</details>
|
|
210
|
-
|
|
211
|
-
<details>
|
|
212
|
-
<summary><strong>🐍 Python</strong> — 3 skills</summary>
|
|
213
|
-
|
|
214
|
-
| Skill | Description |
|
|
215
|
-
|-------|-------------|
|
|
216
|
-
| `python-pro` | Master Python 3.12+ with modern features, async programming, performance optimization, and production-ready practices. uv, ruff, pydantic, and FastAPI. |
|
|
217
|
-
| `python-patterns` | Python development principles. Framework selection, async patterns, type hints, project structure. |
|
|
218
|
-
| `python-performance-optimization` | Profile and optimize Python code using cProfile, memory profilers, and performance best practices. |
|
|
219
|
-
|
|
220
|
-
</details>
|
|
221
|
-
|
|
222
|
-
<details>
|
|
223
|
-
<summary><strong>🦕 TypeScript & JavaScript</strong> — 2 skills</summary>
|
|
224
|
-
|
|
225
|
-
| Skill | Description |
|
|
226
|
-
|-------|-------------|
|
|
227
|
-
| `typescript-pro` | Master TypeScript with advanced types, generics, and strict type safety. Complex type systems, decorators, and enterprise-grade patterns. |
|
|
228
|
-
| `modern-javascript-patterns` | Mastering modern JavaScript (ES6+) features, functional programming patterns, and best practices. |
|
|
229
|
-
|
|
230
|
-
</details>
|
|
231
|
-
|
|
232
|
-
<details>
|
|
233
|
-
<summary><strong>🧠 AI, LLM & Prompt Engineering</strong> — 13 skills</summary>
|
|
234
|
-
|
|
235
|
-
| Skill | Description |
|
|
236
|
-
|-------|-------------|
|
|
237
|
-
| `ai-product` | Build AI-powered products right — from demo to production. |
|
|
238
|
-
| `gemini-api-dev` | Access Google's most advanced AI models via the Gemini API. |
|
|
239
|
-
| `gemini-api-integration` | Integrating Google Gemini API: model selection, multimodal inputs, streaming, function calling. |
|
|
240
|
-
| `llm-app-patterns` | Production-ready patterns for building LLM applications. |
|
|
241
|
-
| `llm-application-dev-ai-assistant` | AI assistant development: intelligent conversational interfaces, chatbots, and AI-powered applications. |
|
|
242
|
-
| `llm-prompt-optimizer` | Applies proven prompt engineering techniques to boost output quality, reduce hallucinations, and cut token usage. |
|
|
243
|
-
| `llm-structured-output` | Reliable JSON, enums, and typed objects from LLMs using response_format, tool_use, and schema-constrained decoding. |
|
|
244
|
-
| `local-llm-expert` | Local LLM inference, model selection, VRAM optimization. Ollama, llama.cpp, vLLM, and LM Studio. GGUF, EXL2 quantization. |
|
|
245
|
-
| `prompt-engineer` | Transforms prompts using frameworks (RTF, RISEN, Chain of Thought, RODES, Chain of Density, RACE, RISE, STAR, SOAP, CLEAR, GROW). |
|
|
246
|
-
| `prompt-engineering-patterns` | Advanced prompt engineering techniques to maximize LLM performance, reliability, and controllability. |
|
|
247
|
-
| `rag-engineer` | Building Retrieval-Augmented Generation systems. Embedding models, vector databases, chunking, and retrieval optimization. |
|
|
248
|
-
| `rag-implementation` | RAG implementation: embedding selection, vector database setup, chunking strategies, and retrieval optimization. |
|
|
249
|
-
| `vector-database-engineer` | Vector databases, embedding strategies, and semantic search. Pinecone, Weaviate, Qdrant, Milvus, and pgvector. |
|
|
250
|
-
|
|
251
|
-
</details>
|
|
252
|
-
|
|
253
|
-
<details>
|
|
254
|
-
<summary><strong>🤖 Agents & Multi-Agent Systems</strong> — 8 skills</summary>
|
|
255
|
-
|
|
256
|
-
| Skill | Description |
|
|
257
|
-
|-------|-------------|
|
|
258
|
-
| `agent-manager-skill` | Manage multiple local CLI agents via tmux sessions (start/stop/monitor/assign) with cron-friendly scheduling. |
|
|
259
|
-
| `agent-memory-mcp` | Hybrid memory system providing persistent, searchable knowledge management for AI agents. |
|
|
260
|
-
| `agent-orchestrator` | Meta-skill orchestrating all agents in the ecosystem. Automatic skill scanning, capability matching, and multi-skill workflow coordination. |
|
|
261
|
-
| `agent-pipeline` | Reference for the BDB lifecycle (define → plan → build → verify/review → ship) that `/startcycle-graph`'s dispatcher graph actually runs. |
|
|
262
|
-
| `agent-tool-builder` | Tool design from schema to error handling — the difference between an agent that works and one that hallucinates. |
|
|
263
|
-
| `ai-agent-development` | AI agent development workflow: autonomous agents, multi-agent systems, and agent orchestration with CrewAI, LangGraph. |
|
|
264
|
-
| `crewai` | Expert in CrewAI — the leading role-based multi-agent framework. |
|
|
265
|
-
| `subagent-driven-development` | Execute implementation plans with independent tasks in the current session. |
|
|
266
|
-
|
|
267
|
-
</details>
|
|
268
|
-
|
|
269
|
-
<details>
|
|
270
|
-
<summary><strong>🕷️ Web Scraping & Automation</strong> — 4 skills</summary>
|
|
271
|
-
|
|
272
|
-
| Skill | Description |
|
|
273
|
-
|-------|-------------|
|
|
274
|
-
| `apify-lead-generation` | Scrape leads from multiple platforms using Apify Actors. |
|
|
275
|
-
| `apify-ultimate-scraper` | AI-driven data extraction from 55+ Actors across all major platforms. |
|
|
276
|
-
| `browser-automation` | Browser automation for web testing, scraping, and AI agent interactions. Selectors, waiting strategies, anti-detection. |
|
|
277
|
-
| `web-scraper` | Intelligent multi-strategy web scraping. Structured data extraction from tables, lists, prices. Pagination, monitoring, CSV/JSON export. |
|
|
278
|
-
|
|
279
|
-
</details>
|
|
280
|
-
|
|
281
|
-
<details>
|
|
282
|
-
<summary><strong>🔥 Firecrawl Workspace Agents</strong> — 13 skills</summary>
|
|
283
|
-
|
|
284
|
-
| Skill | Description |
|
|
285
|
-
|-------|-------------|
|
|
286
|
-
| `firecrawl` | Search, scrape, and interact with the web via the Firecrawl CLI. Real-time web search with full page content. |
|
|
287
|
-
| `firecrawl-agent` | AI-powered autonomous data extraction that navigates complex sites and returns structured JSON. |
|
|
288
|
-
| `firecrawl-build` | Integrate Firecrawl into product code for web scraping, crawling, searching, and interaction. |
|
|
289
|
-
| `firecrawl-build-interact` | Integrate Firecrawl `/interact` for dynamic pages and browser actions after scraping. |
|
|
290
|
-
| `firecrawl-build-onboarding` | Get Firecrawl credentials and SDK setup into a project. |
|
|
291
|
-
| `firecrawl-build-scrape` | Integrate Firecrawl `/scrape` for single-page extraction in product code. |
|
|
292
|
-
| `firecrawl-build-search` | Integrate Firecrawl `/search` into product code and agent workflows. |
|
|
293
|
-
| `firecrawl-crawl` | Bulk extract content from an entire website or site section. Depth limits, path filtering, concurrent extraction. |
|
|
294
|
-
| `firecrawl-download` | Download an entire website as local files — markdown, screenshots, or multiple formats per page. |
|
|
295
|
-
| `firecrawl-interact` | Control a live browser session: click buttons, fill forms, navigate flows, extract data. |
|
|
296
|
-
| `firecrawl-map` | Discover and list all URLs on a website with optional search filtering. |
|
|
297
|
-
| `firecrawl-scrape` | Extract clean markdown from any URL, including JavaScript-rendered SPAs. |
|
|
298
|
-
| `firecrawl-search` | Web search with full page content extraction. Real search results with optional full-page markdown. |
|
|
299
|
-
|
|
300
|
-
</details>
|
|
301
|
-
|
|
302
|
-
<details>
|
|
303
|
-
<summary><strong>📈 SEO & Marketing</strong> — 7 skills</summary>
|
|
304
|
-
|
|
305
|
-
| Skill | Description |
|
|
306
|
-
|-------|-------------|
|
|
307
|
-
| `copywriting` | Rigorous, conversion-focused marketing copy for landing pages and emails. |
|
|
308
|
-
| `geo-fundamentals` | Generative Engine Optimization for AI search engines (ChatGPT, Claude, Perplexity). |
|
|
309
|
-
| `programmatic-seo` | Design and evaluate programmatic SEO strategies at scale using templates and structured data. |
|
|
310
|
-
| `schema-markup` | Design, validate, and optimize schema.org structured data for SEO impact. |
|
|
311
|
-
| `seo` | Broad SEO audit: technical SEO, on-page SEO, schema, sitemaps, content quality, AI search readiness, and GEO. |
|
|
312
|
-
| `seo-audit` | Diagnose and audit SEO issues affecting crawlability, indexation, rankings, and organic performance. |
|
|
313
|
-
| `seo-technical` | Audit technical SEO: crawlability, indexability, security, URLs, mobile, Core Web Vitals, structured data, JS rendering. |
|
|
314
|
-
|
|
315
|
-
</details>
|
|
316
|
-
|
|
317
|
-
<details>
|
|
318
|
-
<summary><strong>🚀 DevOps, Infrastructure & CLI</strong> — 12 skills</summary>
|
|
319
|
-
|
|
320
|
-
| Skill | Description |
|
|
321
|
-
|-------|-------------|
|
|
322
|
-
| `bash-linux` | Bash/Linux terminal patterns. Critical commands, piping, error handling, scripting. |
|
|
323
|
-
| `cloudflare-workers-expert` | Cloudflare Workers and Edge Computing. Wrangler, KV, D1, Durable Objects, and R2 storage. |
|
|
324
|
-
| `docker-expert` | Advanced Docker containerization: optimization, security hardening, multi-stage builds, orchestration, and production deployment. |
|
|
325
|
-
| `github` | Use the `gh` CLI for issues, pull requests, Actions runs, and GitHub API queries. |
|
|
326
|
-
| `github-actions-templates` | Production-ready GitHub Actions workflow patterns for testing, building, and deploying. |
|
|
327
|
-
| `github-repo` | Standards and workflows for writing, formatting, sanitizing, and publishing high-quality GitHub repositories. |
|
|
328
|
-
| `github-workflow-automation` | Patterns for automating GitHub workflows with AI assistance. |
|
|
329
|
-
| `monorepo-management` | Efficient, scalable monorepos enabling code sharing, consistent tooling, and atomic changes. |
|
|
330
|
-
| `os-scripting` | OS and shell scripting troubleshooting for Linux, macOS, and Windows. |
|
|
331
|
-
| `posix-shell-pro` | Strict POSIX sh scripting for maximum portability across Unix-like systems. |
|
|
332
|
-
| `tmux` | tmux session, window, and pane management for terminal multiplexing and persistent remote workflows. |
|
|
333
|
-
| `turborepo-caching` | Configure Turborepo for efficient monorepo builds with local and remote caching. |
|
|
334
|
-
|
|
335
|
-
</details>
|
|
336
|
-
|
|
337
|
-
<details>
|
|
338
|
-
<summary><strong>🧊 3D, Motion & Video</strong> — 3 skills</summary>
|
|
339
|
-
|
|
340
|
-
| Skill | Description |
|
|
341
|
-
|-------|-------------|
|
|
342
|
-
| `bdbmediastorm` | Master ideation and brainstorming engine for live shows, show-control, and event tech. Multi-agent planning for hardware constraints, signal routing, protocols (OSC, Art-Net, DMX, MIDI, SMPTE), TouchDesigner, Resolume, and grandMA3. |
|
|
343
|
-
| `remotion` | Generate walkthrough videos from Stitch projects using Remotion with smooth transitions, zooming, and text overlays. |
|
|
344
|
-
| `threejs-skills` | Create 3D scenes, interactive experiences, and visual effects using Three.js and WebGL. |
|
|
345
|
-
|
|
346
|
-
</details>
|
|
347
|
-
|
|
348
|
-
<details>
|
|
349
|
-
<summary><strong>📝 Documentation, Planning & Quality</strong> — 19 skills</summary>
|
|
350
|
-
|
|
351
|
-
| Skill | Description |
|
|
352
|
-
|-------|-------------|
|
|
353
|
-
| `architect-review` | Master software architect specializing in modern architecture review. |
|
|
354
|
-
| `clean-code` | Principles of "Clean Code" by Robert C. Martin. Transform "code that works" into "code that is clean." |
|
|
355
|
-
| `concise-planning` | Generate clear, actionable, and atomic checklists for coding tasks. |
|
|
356
|
-
| `debugger` | Debugging specialist for errors, test failures, and unexpected behavior. |
|
|
357
|
-
| `deep-research` | Autonomous research tasks that plan, search, read, and synthesize comprehensive reports. |
|
|
358
|
-
| `documentation` | Documentation generation: API docs, architecture docs, README files, code comments, and technical writing. |
|
|
359
|
-
| `executing-plans` | Execute written implementation plans in a separate session with review checkpoints. |
|
|
360
|
-
| `git-advanced-workflows` | Advanced Git techniques: clean history, effective collaboration, and recovery. |
|
|
361
|
-
| `git-pr-review` | Generate concise and structured PR descriptions from commit history. |
|
|
362
|
-
| `linear-claude-skill` | Manage Linear issues, projects, and teams. |
|
|
363
|
-
| `planning-with-files` | Use persistent markdown files as "working memory on disk." |
|
|
364
|
-
| `product-manager-toolkit` | Essential tools and frameworks for modern product management, from discovery to delivery. |
|
|
365
|
-
| `readme` | Expert technical writer creating absurdly thorough project documentation. |
|
|
366
|
-
| `simplify-code` | Review a diff for clarity and safe simplifications, then optionally apply low-risk fixes. |
|
|
367
|
-
| `software-architecture` | Quality-focused software architecture for code design, analysis, and development. |
|
|
368
|
-
| `systematic-debugging` | Use before proposing fixes when encountering any bug, test failure, or unexpected behavior. |
|
|
369
|
-
| `tdd-workflow` | Test-Driven Development workflow principles. RED-GREEN-REFACTOR cycle. |
|
|
370
|
-
| `test-driven-development` | Use when implementing any feature or bugfix, before writing implementation code. |
|
|
371
|
-
| `writing-plans` | Use when you have a spec or requirements for a multi-step task, before touching code. |
|
|
372
|
-
|
|
373
|
-
</details>
|
|
374
|
-
|
|
375
|
-
<details>
|
|
376
|
-
<summary><strong>📦 Utilities & Integrations</strong> — 15 skills</summary>
|
|
377
|
-
|
|
378
|
-
| Skill | Description |
|
|
379
|
-
|-------|-------------|
|
|
380
|
-
| `mcp-manage` | Manages the BDB specialized MCP servers including Unreal Engine, Rhino 7/8, DaVinci Resolve, grandMA3, Resolume, GitHub, Chrome DevTools, and TouchDesigner. |
|
|
381
|
-
| `n8n-code-javascript` | Write JavaScript code in n8n Code nodes. `$input`/`$json`/`$node` syntax, HTTP requests, DateTime handling. |
|
|
382
|
-
| `n8n-code-python` | Write Python code in n8n Code nodes. `_input`/`_json`/`_node` syntax and standard library. |
|
|
383
|
-
| `n8n-expression-syntax` | Validate n8n expression syntax and fix common errors. `{{}}` syntax, `$json`/`$node` variables. |
|
|
384
|
-
| `n8n-mcp-tools-expert` | Expert guide for using n8n-mcp MCP tools effectively. Tool selection, parameter formats, common patterns. |
|
|
385
|
-
| `n8n-workflow-patterns` | Proven architectural patterns for building n8n workflows. |
|
|
386
|
-
| `notion-automation` | Automate Notion tasks via Rube MCP (Composio): pages, databases, blocks, comments, users. |
|
|
387
|
-
| `obsidian-markdown` | Obsidian Flavored Markdown with wikilinks, embeds, callouts, properties, and Obsidian-specific syntax. |
|
|
388
|
-
| `playwright-skill` | Browser automation and testing with Playwright. Path-aware installation across plugin systems. |
|
|
389
|
-
| `slack-automation` | Automate Slack workspace operations: messaging, search, channel management, and reaction workflows. |
|
|
390
|
-
| `token-saver-config` | Context window output compression engine for CLI commands (60-99% token reduction). |
|
|
391
|
-
| `webapp-testing` | Test local web applications with native Python Playwright scripts. |
|
|
392
|
-
| `web-performance-optimization` | Optimize website performance: loading speed, Core Web Vitals, bundle size, caching, and runtime performance. |
|
|
393
|
-
|
|
394
|
-
</details>
|
|
395
|
-
|
|
396
|
-
<details>
|
|
397
|
-
<summary><strong>🌀 BDB Ecosystem & Methodologies</strong> — 4 skills</summary>
|
|
398
|
-
|
|
399
|
-
| Skill | Description |
|
|
400
|
-
|-------|-------------|
|
|
401
|
-
| `bdbrainstorm` | Combines multi-agent brainstorming, the `/grill-me` slash command, subagent-driven-development, and ui-ux-pro-max for comprehensive ideation. |
|
|
402
|
-
| `memb-skill` | BDB local-first long-term memory engine (memB). Query, remember, and adapt preferences, code architectures, and developer patterns across tasks. |
|
|
403
|
-
| `memb-ingest` | Deep scan and ingest project files (.md, .json, AGENTS.md, .openwiki) and past conversation logs into the local memB vector memory engine. |
|
|
404
|
-
| `openwiki-skill` | Direct Gemini-native integration of OpenWiki for autonomous, high-agency documentation management and release notes maintenance. |
|
|
405
|
-
|
|
406
|
-
</details>
|
|
407
|
-
|
|
408
|
-
## 🔄 AOS: The Dispatcher Graph
|
|
409
|
-
|
|
410
|
-
v4.0.0 replaces the old linear "5-agent pipeline" prose with a real, runnable
|
|
411
|
-
state machine. The contract lives in [`.agents/graph.md`](.agents/graph.md),
|
|
412
|
-
the node roster in [`.agents/nodes.json`](.agents/nodes.json), and the
|
|
413
|
-
executable dispatcher in
|
|
414
|
-
[`.claude/workflows/startcycle-dispatch.mjs`](.claude/workflows/startcycle-dispatch.mjs).
|
|
415
|
-
|
|
416
|
-
**The one rule everything else follows: nodes never invoke each other.** A
|
|
417
|
-
single dispatcher reads `production_artifacts/state.json` after each node
|
|
418
|
-
returns and decides what runs next. There is no hand-off chain, no agent
|
|
419
|
-
telling another agent to go — which is what stops the prompt drift that makes
|
|
420
|
-
long agent pipelines wander off course.
|
|
87
|
+
```bash
|
|
88
|
+
npx -y @hybridlabor-api/aos@latest -y --platforms=2 # Claude only, all defaults
|
|
89
|
+
npx -y @hybridlabor-api/aos@latest -y --platforms=1,5 --mcps=none
|
|
90
|
+
npx -y @hybridlabor-api/aos@latest --dry-run # print what would change
|
|
91
|
+
```
|
|
421
92
|
|
|
422
|
-
|
|
423
|
-
flowchart LR
|
|
424
|
-
U(["👤 User"])
|
|
425
|
-
A["<b>Architect</b><br/><span>System Plan (00)</span>"]
|
|
426
|
-
T["<b>TechLead</b><br/><span>Capability Approval</span>"]
|
|
427
|
-
UX["<b>Godmode_UI_UX</b><br/><span>Frontend Spec (01)</span>"]
|
|
428
|
-
EN["<b>Godmode_Engineering</b><br/><span>Backend Schema (02)</span>"]
|
|
429
|
-
ME["<b>Godmode_Media</b><br/><span>EventTech (03)</span>"]
|
|
430
|
-
R["<b>Reviewer</b><br/><span>Doubt-Driven QA</span>"]
|
|
431
|
-
S["<b>Shipping</b><br/><span>GO-Gate</span>"]
|
|
93
|
+
`--platforms=` values: `0` universal (all detected), `1` Antigravity, `2` Claude Desktop / Claude Code, `3` Cursor, `5` Codex CLI, `6` Windsurf, `7` Roo Code / Cline, `8` Aider, `10` AOS CLI. `4` (custom paths) needs the interactive menu. `--mcps=<name,name>|all|none` picks the MCP subset. `--verbose` and `--no-intro` do what they say.
|
|
432
94
|
|
|
433
|
-
|
|
434
|
-
T --> UX & EN & ME
|
|
435
|
-
UX & EN & ME --> R
|
|
436
|
-
R --> S
|
|
95
|
+
### Local project harness
|
|
437
96
|
|
|
438
|
-
|
|
439
|
-
R -.->|Reviewer Findings · repair loop| UX
|
|
440
|
-
S -.->|gate failed · Shipping names the owner| EN
|
|
441
|
-
R -.->|needs_human · no-progress guard| U
|
|
97
|
+
Instead of installing into `$HOME`, drop only the dispatcher contract (`.agents/`, the gate hooks, the agent definitions, the `/startcycle-graph` workflow, the OpenCode plugin) into one repository:
|
|
442
98
|
|
|
443
|
-
|
|
444
|
-
|
|
445
|
-
class A,T,UX,EN,ME,R,S box
|
|
446
|
-
class U user
|
|
447
|
-
linkStyle 8,9,10,11 stroke:#7d8799,stroke-width:1px,color:#9aa4b8
|
|
99
|
+
```bash
|
|
100
|
+
cd your-project && npx -y @hybridlabor-api/aos@latest --project-harness -y
|
|
448
101
|
```
|
|
449
102
|
|
|
450
|
-
|
|
103
|
+
---
|
|
451
104
|
|
|
452
|
-
|
|
453
|
-
|---|---|
|
|
454
|
-
| **Reviewer isolation** | The Reviewer reads build artifacts and the plan's contract — never the original `goal`, never a build node's reasoning or its claim that the work is done. Passing the implementer's claim biases a reviewer toward agreement; withholding it is what makes the review adversarial rather than a rubber stamp. |
|
|
455
|
-
| **No-progress guard** | If a repair cycle comes back reporting the *same* blocking finding ID as the previous one, nothing is actually being fixed. The run escalates to a human instead of burning iterations re-running an identical loop. |
|
|
456
|
-
| **Per-node state fragments** | Build nodes run in parallel and each writes its own `state.d/<node>.json` fragment, merged afterward — they never write `state.json` directly. Parallel writers to one JSON file is a lost-update race; fragments remove the race by construction. |
|
|
457
|
-
| **Iteration ceiling** | `max_iterations` (default 3) stops the loop unconditionally, deliberately set below Claude Code's own 8-consecutive-Stop-hook override so the run's own escalation message reaches you first. |
|
|
458
|
-
| **In-loop human edge** | Any node can set `needs_human: true` and stop the run. Full autonomy sounds good until a node hits something only a human can decide — this is an explicit edge in the graph, not an interruption of it. |
|
|
459
|
-
| **GO-gated shipping** | Reaching `ready_to_ship` is not shipping. `git push`, `npm publish`, `npm version` and recursive `rm` are blocked by a `PreToolUse` hook ([`.claude/hooks/go-gate.mjs`](.claude/hooks/go-gate.mjs)) unless your immediately preceding message is the literal word **GO**. It is a hook, not a rule an agent reads and tries to follow — it fires before any permission-mode check and cannot be argued around. |
|
|
105
|
+
## Supported harnesses
|
|
460
106
|
|
|
461
|
-
|
|
107
|
+
What the installer writes for each target. Paths are the defaults; the installer only writes to harnesses it actually detects.
|
|
462
108
|
|
|
463
|
-
|
|
|
464
|
-
|
|
465
|
-
|
|
|
466
|
-
|
|
|
467
|
-
|
|
|
109
|
+
| Harness | Skills | Subagents | Hooks | Plugin / rules |
|
|
110
|
+
|---|---|---|---|---|
|
|
111
|
+
| Claude Code / Claude Desktop | `~/.claude/skills` | `~/.claude/agents` | `~/.claude/hooks` + `settings.json` (GO gate, graph gate, env-file protection, Conventional Commits, memB inject, trail relay) | `.claude-plugin/` manifest ships in the repo (see Contributing) |
|
|
112
|
+
| Google Antigravity | `~/.gemini/config/skills` | `~/.gemini/config/agents` | `~/.gemini/config/hooks.json` and `~/.gemini/antigravity-cli/hooks.json` | — |
|
|
113
|
+
| Codex CLI | `~/.codex/skills` | `~/.codex/agents` | `~/.codex/hooks` + `config.toml` | `.codex-plugin/` |
|
|
114
|
+
| OpenCode | `~/.config/opencode/skills` | `~/.opencode/agents` | via plugin | `bdb-aos.js` plugin + `/startcycle-graph` command, registered in `opencode.jsonc`; keeps a `/startcycle-graph` run moving on `session.idle` |
|
|
115
|
+
| Cursor | `~/.cursor/skills` | — | — | `.cursor/rules` (project) |
|
|
116
|
+
| Windsurf | `~/.windsurf/bdb-skills` | — | — | `mcp.json` |
|
|
117
|
+
| Roo Code / Cline | `~/.roo/skills` | — | — | `.roomodes` (project) |
|
|
118
|
+
| Aider | `~/.aider/bdb-skills` | — | — | — |
|
|
119
|
+
| AOS CLI (`pi`) | reads `~/.agents/skills` | `~/.agents/AGENTS.md` as system prompt | — | no MCP; separate install, Node >= 22.19 — see [packages/aos-cli](packages/aos-cli/README.md) |
|
|
468
120
|
|
|
469
|
-
|
|
470
|
-
whether Antigravity, OpenCode or Codex are present and falls back to Claude
|
|
471
|
-
Code's own subagents when none are. Model tiers are forced per role rather than
|
|
472
|
-
inherited from your session, so a mechanical worker step doesn't silently run
|
|
473
|
-
on Opus because that's what you happened to have selected.
|
|
121
|
+
Every install also writes the universal copy to `~/.agents/skills`, which is what the AOS CLI and the `skills` CLI read.
|
|
474
122
|
|
|
475
123
|
---
|
|
476
124
|
|
|
477
|
-
##
|
|
125
|
+
## The pipelines
|
|
478
126
|
|
|
479
|
-
|
|
127
|
+
The contract lives in [`.agents/graph.md`](.agents/graph.md), the node roster in [`.agents/nodes.json`](.agents/nodes.json), the executable dispatcher in [`.claude/workflows/startcycle-dispatch.mjs`](.claude/workflows/startcycle-dispatch.mjs).
|
|
480
128
|
|
|
481
|
-
|
|
129
|
+
**One rule: nodes never invoke each other.** A dispatcher reads `production_artifacts/state.json` after each node returns and decides what runs next. There is no hand-off chain and no agent telling another agent to go. This design ensures context fidelity, single-place auditability of routing logic, and portability across harnesses.
|
|
482
130
|
|
|
483
|
-
|
|
131
|
+
Each node reads the plan and its own prior state, executes its work, writes its artifact and state fragments, and returns. The dispatcher merges per-node state fragments (`state.d/<node>.json`), evaluates edge predicates, and routes to the next node — or escalates to the user if a no-progress guard triggers (same blocking finding on the second repair cycle) or the iteration ceiling is reached.
|
|
132
|
+
|
|
133
|
+
```mermaid
|
|
134
|
+
flowchart LR
|
|
135
|
+
U(["User"]) --> A["Architect"] --> T["TechLead"]
|
|
136
|
+
T --> UX["Godmode_UI_UX"] & EN["Godmode_Engineering"] & ME["Godmode_Media"]
|
|
137
|
+
UX & EN & ME --> R["Reviewer"] --> S["Shipping"]
|
|
138
|
+
T -.->|reject| A
|
|
139
|
+
R -.->|findings| UX
|
|
140
|
+
R -.->|needs_human| U
|
|
141
|
+
```
|
|
484
142
|
|
|
485
|
-
|
|
143
|
+
| Command | Machinery | Use when |
|
|
144
|
+
|---|---|---|
|
|
145
|
+
| `/startcycle` | Linear chain, file hand-offs in `production_artifacts/`. No state machine, no repair loop. `/startcycle --skill=<name> <goal>` forces a skill into every node. | A straightforward build with the agent roster but without the ceremony. |
|
|
146
|
+
| `/startcycle-graph` | The full graph: durable `state.json`, Reviewer repair loop, automated quality gate, human escalation. | Feature work where correctness matters more than speed and you want an audit trail. |
|
|
147
|
+
| `/startcycle-graph-user` | Throwaway 2–4 node fan-out. Nothing persistent. Model-tiered per role; uses Antigravity, OpenCode or Codex if installed, Claude Code subagents otherwise. | A one-off "spawn a few workers" in any project. |
|
|
486
148
|
|
|
487
|
-
|
|
149
|
+
What keeps the full graph honest:
|
|
488
150
|
|
|
489
|
-
|
|
151
|
+
| Mechanism | What it prevents |
|
|
152
|
+
|---|---|
|
|
153
|
+
| Reviewer isolation | The Reviewer reads artifacts and the plan's contract, never the build node's reasoning or its claim that it is done. |
|
|
154
|
+
| No-progress guard | A repair cycle that reports the same blocking finding ID as the previous one escalates to a human instead of burning iterations. |
|
|
155
|
+
| Per-node state fragments | Parallel build nodes write `state.d/<node>.json`; the dispatcher merges. No lost-update race on one file. |
|
|
156
|
+
| Iteration ceiling | `max_iterations` (default 3) stops the loop unconditionally. |
|
|
157
|
+
| In-loop human edge | Any node can set `needs_human: true` and stop the run. |
|
|
490
158
|
|
|
491
|
-
|
|
492
|
-
<summary><strong>🎨 Adobe Creative Cloud (Illustrator, Photoshop, After Effects, Premiere Pro)</strong></summary>
|
|
159
|
+
## The GO gate
|
|
493
160
|
|
|
494
|
-
|
|
495
|
-
- **Direct OS-Native Bridge (`bdb_adobe_mcp`)**: Runs zero-install scripting.
|
|
496
|
-
- **macOS:** Targets application bundle IDs directly via AppleScript `do javascript` / `DoScript` command streams.
|
|
497
|
-
- **Windows:** Automatically queries and instantiates local COM objects via PowerShell wrapper scripts and executes transient `.jsx` ExtendScript code.
|
|
498
|
-
- **Cross-Platform UXP WebSocket Bridge (`bdb_adobe_uxp_mcp`)**: A three-tier WebSocket proxy (Node.js server on port 8080 + native UXP developer plugins) for deep DOM manipulation and persistent WebSocket sessions inside Photoshop and Premiere Pro, running identically on Windows and macOS.
|
|
499
|
-
</details>
|
|
161
|
+
Reaching `ready_to_ship` is not shipping. [`.claude/hooks/go-gate.mjs`](.claude/hooks/go-gate.mjs) is a `PreToolUse` hook that blocks `git push`, `npm publish`, `npm version` and recursive `rm` unless your immediately preceding message is the literal word **GO**. It is a hook, not a rule an agent is asked to respect: it fires before any permission-mode check and cannot be argued around. The installer wires the same gate into Antigravity, Codex and OpenCode; on harnesses without hook support the rule in [AGENTS.md](AGENTS.md) applies and the agent is the enforcement. A subagent never inherits its orchestrator's GO, and a failed release command needs a fresh one.
|
|
500
162
|
|
|
501
|
-
|
|
502
|
-
<summary><strong>🎬 DaVinci Resolve (Triple Coverage)</strong></summary>
|
|
163
|
+
---
|
|
503
164
|
|
|
504
|
-
|
|
505
|
-
- **Studio: `bdb_davinci_mcp_studio`**: The official Node.js server (wrapping samuelgursky) for advanced direct timeline and project management in Resolve Studio.
|
|
506
|
-
- **Fallback: `bdb_davinci_mcp_fallback`**: Hoyt-harness professional python server for Studio scripting.
|
|
507
|
-
</details>
|
|
165
|
+
## Tools
|
|
508
166
|
|
|
509
|
-
|
|
510
|
-
<summary><strong>📐 Rhino 3D & Grasshopper (Twin-Engine)</strong></summary>
|
|
167
|
+
|
|
511
168
|
|
|
512
|
-
|
|
513
|
-
- **Fallback: `bdb_rhino_mcp_fallback`**: The GOLEM 3D server with 105 tools to dynamically manipulate Rhino 8 assets, execute scripts, and solve Grasshopper definitions.
|
|
514
|
-
</details>
|
|
169
|
+
*Conceptual overview from v3.4.0. Components have grown since; the sections below are current.*
|
|
515
170
|
|
|
516
|
-
|
|
517
|
-
|
|
171
|
+
| Tool | Command | What it does |
|
|
172
|
+
|---|---|---|
|
|
173
|
+
| Plan Canvas | `aos-plan-canvas open <file>` (skill `plan-canvas`) | Opens a plan or HTML artifact in a local browser canvas where you annotate elements, chat, and approve or request changes. Plans from the pipelines open here by default. |
|
|
174
|
+
| agenttrail | `aos-trail` (skill `agenttrail`, port 5330) | Live board of a multi-agent build: which component, which agent or harness, what is done, what is stuck. Fed by the trail-relay hooks and by `mcsc`. |
|
|
175
|
+
| archify | `aos-archify` (skill `archify`) | Validated architecture, sequence, data-flow and state diagrams as standalone HTML with SVG export; accepts Mermaid. |
|
|
176
|
+
| AOS Store | `aos store list \| search <q> \| install <name> [--project]`, `aos store ui` (also `aos-store`, slash command `/aos-store`) | Browse and install AOS Core and ECC skills and agents. The web UI on `http://127.0.0.1:4322` shows what is installed, previews the exact target paths, and installs only after you confirm. Multi-file skills are installed completely. `list` and `search` read a pinned offline index. |
|
|
177
|
+
| Launchpad | `aos-dashboard [--port 7900] [--no-open]` | One page with every local BDB service (memB, Synapse, OpenWiki, AO, Remote, AOS Store): status, start/stop, logs. Registered as an autostart entry. |
|
|
178
|
+
| Doctor | `aos doctor [--json] [--net]` (also `aos-doctor`) | Verifies dependencies, skill placement per harness, daemons, hooks and modules; exits 1 when something needs attention. The first thing to run when anything misbehaves. |
|
|
179
|
+
| Config | `aos-config show \| propose \| set <key> <value>` | Machine-level `~/.agents/aos-config.json`: workspace root, domains, user id. |
|
|
180
|
+
| mcsc | MCP tools `delegate_agy`, `delegate_opencode`, `delegate_codex`, `delegate_smart` (skill `mcsc`) | Delegates a task to another installed CLI harness and streams its tool calls to agenttrail. Preferred over shelling out to the CLI. |
|
|
518
181
|
|
|
519
|
-
|
|
520
|
-
- **Primary: `bdb_vectorworks_mcp`**: Semantic RAG-based search index over VectorScript and Vectorworks API documentation (port 8765) for automated CAD drafting.
|
|
182
|
+
---
|
|
521
183
|
|
|
522
|
-
|
|
523
|
-
- **Primary: `bdb_unreal_mcp`**: Connects via the Unreal Engine 5 Web Remote Control API (port 30010) and the `gimmeDG` toolset. Allows the agent to query, spawn actors, edit materials, write Blueprints, and automate level/sequencer manipulation.
|
|
184
|
+
## AOS CLI
|
|
524
185
|
|
|
525
|
-
|
|
526
|
-
- **Primary: `bdb_blender_mcp`**: BlenderMCP socket integration for scene layout, mesh generation, and viewport controls.
|
|
527
|
-
- **Fallback: `bdb_blender_mcp_fallback`**: djeada's python server for managing Blender TCP connections and raw python scripting.
|
|
186
|
+
A lightweight CLI harness built on [pi](https://github.com/earendil-works/pi), a coding agent that runs in the terminal. AOS CLI reads `~/.agents/skills` (written by the installer) and `~/.agents/AGENTS.md` (the dispatcher graph as system instructions), runs no MCP servers of its own, and needs **Node >= 22.19** (pi's floor, higher than the main AOS installer).
|
|
528
187
|
|
|
529
|
-
|
|
530
|
-
-
|
|
531
|
-
-
|
|
188
|
+
```bash
|
|
189
|
+
aos-cli "what is the fastest way to fix this bug"
|
|
190
|
+
aos-cli --continue # resume the previous session
|
|
191
|
+
```
|
|
532
192
|
|
|
533
|
-
|
|
534
|
-
- **grandMA3**: `bdb_ma3_mcp` sends OSC/UDP command streams directly to your grandMA3 console (port 8000) to automate cues, macros, and patch fixtures.
|
|
535
|
-
- **Resolume**: `bdb_resolume_mcp` wraps Arena's REST API (port 8080) to sequence layers, query statuses, and trigger clips.
|
|
193
|
+
The CLI launcher (`packages/aos-cli/bin/aos-cli.mjs`) ships with an AOS-themed dark mode (`aos.json`), the ten core skills from `core-skills.json` (ask-tim, aos-setup, systematic-debugging, archify, etc.), and two in-session read-only commands (`/aos` shows the install menu; `/aos-status` runs the health check).
|
|
536
194
|
|
|
537
|
-
|
|
538
|
-
- **macOS/Linux: `zavora_computer_use`**: Bundled with precompiled native Rust NAPI binary objects (macOS arm64/x64, Linux) to control mouse, keyboard, windows, and apps without runtime compile errors.
|
|
539
|
-
- **Windows: `bdb_windows_computer_use`**: Native python-based Win32 / COM / UIAutomation controller with local OCR (Tesseract) support for advanced Windows GUI automation.
|
|
195
|
+
Install it through the AOS installer with the AOS CLI target: `npx -y @hybridlabor-api/aos@latest -y --platforms=10`. The package is private and is not published on npm, so `npm i -g @hybridlabor-api/aos-cli` does not work.
|
|
540
196
|
|
|
541
|
-
### 🧠 Local Semantic Brain (memB)
|
|
542
|
-
- **`memb_mcp`**: Exposes standard long-term memory tools (`add_memory`, `search_memory`, `delete_memory`, `list_memories`) using a completely local, offline-first vector engine (powered by a bundled 30MB ONNX model and SQLite).
|
|
543
|
-
</details>
|
|
544
197
|
|
|
545
198
|
---
|
|
546
199
|
|
|
547
|
-
##
|
|
548
|
-
|
|
549
|
-
|
|
550
|
-
|
|
551
|
-
|
|
552
|
-
|
|
553
|
-
|
|
554
|
-
|
|
555
|
-
|
|
556
|
-
|
|
557
|
-
|
|
558
|
-
- [`bdb-after-effects-mcp`](skills/global_config/bdb-after-effects-mcp/SKILL.md)
|
|
559
|
-
- [`bdb-vectorworks-mcp`](skills/global_config/bdb-vectorworks-mcp/SKILL.md)
|
|
560
|
-
- [`bdb-touchdesigner-mcp`](skills/global_config/bdb-touchdesigner-mcp/SKILL.md)
|
|
561
|
-
- [`bdb-computer-use-mcp`](skills/global_config/bdb-computer-use-mcp/SKILL.md)
|
|
562
|
-
- [`bdb-grandma3-mcp`](skills/global_config/bdb-grandma3-mcp/SKILL.md)
|
|
563
|
-
- [`bdb-resolume-mcp`](skills/global_config/bdb-resolume-mcp/SKILL.md)
|
|
564
|
-
- [`bdb-adobe-suite-mcp`](skills/global_config/bdb-adobe-suite-mcp/SKILL.md)
|
|
565
|
-
- [`bdb-memb-mcp`](skills/global_config/bdb-memb-mcp/SKILL.md)
|
|
566
|
-
- [`openwiki-skill`](skills/global_config/openwiki-skill/SKILL.md): Direct Gemini-native integration of OpenWiki for autonomous, high-agency documentation management and release notes maintenance.
|
|
567
|
-
- [`memb-skill`](skills/global_config/memb-skill/SKILL.md): BDB local-first long-term memory engine (memB). Query, remember, and adapt preferences, code architectures, and developer patterns across tasks.
|
|
568
|
-
</details>
|
|
200
|
+
## Plugins and Marketplace
|
|
201
|
+
|
|
202
|
+
**Plugin manifest:** `.claude-plugin/plugin.json` + `marketplace.json` (generated by `npm run plugin:build`). The Claude marketplace installation path is being finalized; for now, the npm installer above is the supported installation route.
|
|
203
|
+
|
|
204
|
+
**Skills discovery:** Every harness finds skills in its native directory (`~/.claude/skills`, `~/.agents/skills`, `~/.codex/skills`, `~/.roo/skills`, etc.). To browse and install additional skills after install:
|
|
205
|
+
|
|
206
|
+
```bash
|
|
207
|
+
npx skills add hybridlabor-api/aos
|
|
208
|
+
```
|
|
209
|
+
|
|
210
|
+
This discovers all <!-- count:skills -->213<!-- /count --> curated skills and installs them into the universal `~/.agents/skills` directory (used by all harnesses and the AOS CLI).
|
|
569
211
|
|
|
570
212
|
---
|
|
571
213
|
|
|
572
|
-
##
|
|
214
|
+
## Memory and knowledge
|
|
573
215
|
|
|
574
|
-
|
|
216
|
+
Installed as optional modules by the installer; `aos doctor` verifies them and the Launchpad shows them.
|
|
575
217
|
|
|
576
|
-
|
|
577
|
-
- **
|
|
578
|
-
- **
|
|
579
|
-
- **
|
|
580
|
-
- **Changelog & History:** [.openwiki/release_notes.md](.openwiki/release_notes.md)
|
|
581
|
-
- **Code Health Report:** [.openwiki/code_health.md](.openwiki/code_health.md)
|
|
582
|
-
- **Interactive Live Dashboard:** [.openwiki/code_health_dashboard.html](.openwiki/code_health_dashboard.html)
|
|
218
|
+
- **memB** (`@hybridlabor-api/memb`) — local, offline vector memory with an MCP server (`add_memory`, `search_memory`, `list_memories`, `delete_memory`), a WebUI on port 8088, and an ambient hook that injects relevant memories into Claude Code sessions. Skills: `memb-skill`, `memb-ingest`, `bdb-memb-mcp`.
|
|
219
|
+
- **deja** (`@vshulcz/deja-vu`, installed with memB) — indexes your agent transcripts locally with secrets redacted; `deja fix` on an error, `deja wip` when resuming, `deja search` for past sessions. Skill: `deja-memory`.
|
|
220
|
+
- **OpenWiki** (`openwiki` CLI) — generates and refreshes a grounded wiki of a codebase, with a visualizer on port 4321 and a background daemon. Skill: `openwiki-skill`; this repo's own wiki is under [`.openwiki/`](.openwiki/quickstart.md).
|
|
221
|
+
- **Synapse** (`@hybridlabor-api/bdb-synapse`) — renders a repository as a 3D code city and replays agent sessions through it. Skill: `synapse-integration-skill`.
|
|
583
222
|
|
|
584
|
-
|
|
585
|
-
<summary><strong>🧠 Multi-Provider LLM & Zero-Token RepoGraph Architecture</strong></summary>
|
|
223
|
+
`aos-setup` brings a machine to a verified state for all four; `aos-project-init` binds one project to them (slug, wiki, memory, `AGENTS.md`).
|
|
586
224
|
|
|
587
|
-
|
|
588
|
-
- **Google GenAI:** `gemma-4-26b-a4b-it` (default via `google-genai` SDK, with automatic model discovery fallback)
|
|
589
|
-
- **Groq:** `llama-3.3-70b-versatile` (ultra-low latency)
|
|
590
|
-
- **Grok / xAI:** `grok-2-latest`
|
|
591
|
-
- **Nvidia NIM:** `meta/llama-3.3-70b-instruct`
|
|
592
|
-
- **OpenRouter:** `anthropic/claude-3.5-sonnet` (200+ models)
|
|
593
|
-
- **OpenAI:** `gpt-4o-mini` / `gpt-4o`
|
|
594
|
-
- **Offline / Local:** Ollama (`llama3`), LM Studio, or any OpenAI-compatible endpoint.
|
|
595
|
-
2. **RepoGraph Zero-Token Git Analytics:** Analyzes 90-day hotspot velocity, single-author bus factor risk, and maintainability index purely through deterministic local Git analysis—costing **0 LLM tokens**.
|
|
596
|
-
3. **Repowise-Grade Live HTML Dashboard:** `.openwiki/code_health_dashboard.html` provides 6 visual SVG panels (Galaxy Cluster Map, Risk Donut, Bus Factor Matrix, Commit Velocity Churn, Hotspot Leaderboard, Architecture Health Radar) with **60-second live auto-refresh** and integrated memB ADR telemetry.
|
|
597
|
-
</details>
|
|
225
|
+
---
|
|
598
226
|
|
|
599
|
-
|
|
600
|
-
<summary><strong>⚙️ Setting Up the Background Daemon</strong></summary>
|
|
227
|
+
## What's Included
|
|
601
228
|
|
|
602
|
-
|
|
229
|
+
### The <!-- count:agents -->13<!-- /count --> Subagents
|
|
603
230
|
|
|
604
|
-
|
|
605
|
-
```bash
|
|
606
|
-
bash ~/.gemini/config/skills/openwiki-skill/scripts/install_daemon.sh
|
|
607
|
-
```
|
|
231
|
+
The dispatcher graph compiles these agents, available as Claude Code subagents and loadable into Antigravity, Cursor, Codex, OpenCode and others:
|
|
608
232
|
|
|
609
|
-
|
|
610
|
-
|
|
611
|
-
|
|
612
|
-
|
|
233
|
+
| Agent | Purpose |
|
|
234
|
+
|---|---|
|
|
235
|
+
| **Architect** | Turns the user's goal into a system plan. Reads existing architecture before proposing changes. |
|
|
236
|
+
| **TechLead** | Reviews the plan for a capability map (module boundaries, dependency direction, build order) before any build node starts. Approves or rejects back to Architect. |
|
|
237
|
+
| **UI_UX** | Lead Frontend Designer. Enforces Anti-Slop principles, DTCG design tokens, high-agency frontend taste, and fluid motion dynamics. |
|
|
238
|
+
| **Engineering** | Senior Fullstack & Backend Engineer. Enforces Domain-Driven Design, Clean Architecture, TDD cycles, and database best practices. |
|
|
239
|
+
| **Media_EventTech** | Creative-Tech & Show-Control Specialist. Governs 3D modeling, TouchDesigner networks, DaVinci Resolve, lighting, and Resolume. |
|
|
240
|
+
| **Reviewer** | Adversarial review of build-node output against the plan's contract. Modeled on doubt-driven-development discipline. |
|
|
241
|
+
| **Shipping** | Release Gatekeeper & QA Auditor. Runs the automated quality gate (lint, typecheck, tests, a11y, seo) and enforces the GO gate. |
|
|
242
|
+
| **Database Reviewer** | PostgreSQL specialist for query optimization, schema design, security, and performance. |
|
|
243
|
+
| **Security Reviewer** | Security vulnerability detection and remediation. Flags secrets, SSRF, injection, unsafe crypto, and OWASP Top 10. |
|
|
244
|
+
| **Silent-Failure Hunter** | Reviews code for silent failures, swallowed errors, bad fallbacks, and missing error propagation. |
|
|
245
|
+
| **Go-Build Resolver** | Resolves Go build, vet, and compilation errors with minimal changes. |
|
|
246
|
+
| **Opensource Forker** | Forks a project for open-sourcing — strips secrets, replaces internal references, generates `.env.example`. |
|
|
247
|
+
| **Opensource Sanitizer** | Verifies an open-source fork is fully sanitized. Scans for leaked secrets, PII, internal references. |
|
|
248
|
+
|
|
249
|
+
### Skills by Category
|
|
250
|
+
|
|
251
|
+
<!-- count:skills -->213<!-- /count --> curated skills, discoverable by every harness:
|
|
252
|
+
|
|
253
|
+
- **bdb-core** (30 skills): Core AOS infrastructure, pipelines, tools, and utilities — `startcycle`, `startcycle-graph`, `startcycle-graph-user`, `agent-orchestrator`, `agenttrail`, `plan-canvas`, `aos-doctor`, `aos-store`, `bdb-dev-os-skill`, and more.
|
|
254
|
+
- **design-ui-ux** (19 skills): Frontend, UI design, accessibility, tokens, motion, anti-slop — `senior-frontend`, `ui-component`, `ui-review`, `tailwind-patterns`, `shadcn`, `wcag-audit-patterns`, and more.
|
|
255
|
+
- **engineering-method** (46 skills): Architecture, testing, debugging, CI/CD, code quality — `software-architecture`, `test-driven-development`, `systematic-debugging`, `ci-pipeline`, `github-actions-generator`, `dockerfile-validator`, and more.
|
|
256
|
+
- **library** (98 skills): Language/framework specifics — TypeScript, Node.js, Python, React, Postgres, Prisma, Next.js, Drizzle ORM, Go, and more.
|
|
257
|
+
- **media-eventtech** (19 skills): 3D, video, show control, spatial design — `godmode-eventtech`, `synapse-integration-skill`, `threejs-skills`, `blender-expert`, and more.
|
|
258
|
+
- **engineering-hardware** (1 skill): PCB and electrical design — `godmode-hardware-pcb`.
|
|
259
|
+
|
|
260
|
+
The full catalog with detailed descriptions: [docs/skills_table.md](docs/skills_table.md) — note: this file is out of date and lists 164 of 213 skills.
|
|
613
261
|
|
|
614
|
-
|
|
615
|
-
|
|
616
|
-
|
|
617
|
-
|
|
618
|
-
|
|
619
|
-
|
|
620
|
-
|
|
621
|
-
|
|
622
|
-
|
|
623
|
-
|
|
624
|
-
|
|
262
|
+
---
|
|
263
|
+
|
|
264
|
+
## Skills
|
|
265
|
+
|
|
266
|
+
<!-- count:skills -->213<!-- /count --> skills, curated from open-source and proprietary collections, covering the full software development and creative pipeline. Every skill is a directory with a `SKILL.md` frontmatter declaring `name`, `description`, and one `category`: `bdb-core`, `design-ui-ux`, `engineering-method`, `engineering-hardware`, `media-eventtech`, `library`.
|
|
267
|
+
|
|
268
|
+
**Persona Layer:** The **Godmode** skills are specialized personas that directly map to the build and ship nodes of the dispatcher graph:
|
|
269
|
+
|
|
270
|
+
| Godmode | Owns | Maps to |
|
|
271
|
+
|---|---|---|
|
|
272
|
+
| `godmode-engineering` | Domain-Driven Design, Clean Architecture, strict TypeScript/Python, systematic debugging, database best practices. | **Engineering** node |
|
|
273
|
+
| `godmode-ui-ux` | Anti-slop frontend principles, DTCG design tokens, motion dynamics, accessibility (WCAG), high-agency taste. | **UI_UX** node |
|
|
274
|
+
| `godmode-shipping` | Pre-launch checks, automated quality gates, safe rollback procedures, Go-gate enforcement. | **Shipping** node |
|
|
275
|
+
| `godmode-eventtech` | Show control, signal flow, DMX lighting, TouchDesigner networks, Resolume media servers, live-event hardware. | **Media_EventTech** node |
|
|
276
|
+
| `godmode-3d-creation` | MCP-first 3D generation, mesh reconstruction, parametric CAD, spatial modeling. | Optional specialist |
|
|
277
|
+
| `godmode-media-creation` | Video production, timeline assembly, motion design pipelines, OpenMontage, Remotion. | Optional specialist |
|
|
278
|
+
| `godmode-hardware-pcb` | Electrical schematics, PCB layout and routing, KiCad ERC/DRC/DFM gate, enclosure co-design, OpenSCAD. | Optional specialist |
|
|
279
|
+
|
|
280
|
+
**Entry Points & Navigation:**
|
|
281
|
+
- **`ask-tim`** — Skill recommendation by description
|
|
282
|
+
- **`bdbrainstorm`** and **`bdbmediastorm`** — Multi-agent ideation sessions ending in an executable plan
|
|
283
|
+
- **`teamwork-preview`** — Prompt crafting, role delegation, collaboration setup
|
|
284
|
+
- **Grilling family** — `grill-me` (general audit), `grill-with-docs` (documentation-grounded), `triage` (prioritization)
|
|
285
|
+
- **CI/CD & Generators** — `ci-pipeline`, `github-actions-generator`, `dockerfile-generator`, `makefile-generator`
|
|
286
|
+
- **Code Quality** — `bdb-security-audit`, `systematic-debugging`, `silent-failure-hunter`, `bdbresilience`
|
|
287
|
+
- **Framework Specialists** — Full coverage of TypeScript, React, Next.js, Drizzle ORM, Prisma, Python, Go, and more
|
|
288
|
+
|
|
289
|
+
The full catalog with descriptions and details: [docs/skills_table.md](docs/skills_table.md) (note: currently lists 164 of 213).
|
|
290
|
+
|
|
291
|
+
The library is also readable by the `skills` CLI:
|
|
625
292
|
|
|
626
|
-
**Monitor Execution:** Tail the active logs to inspect background documentation rebuild status:
|
|
627
293
|
```bash
|
|
628
|
-
|
|
294
|
+
npx skills add hybridlabor-api/aos
|
|
629
295
|
```
|
|
630
|
-
</details>
|
|
631
296
|
|
|
632
297
|
---
|
|
633
298
|
|
|
634
|
-
##
|
|
635
|
-
|
|
636
|
-
[](https://github.com/hybridlabor-api/bdb-agent-orchestrator)
|
|
637
|
-
[](https://github.com/hybridlabor-api/bdb-agent-orchestrator)
|
|
638
|
-
[](https://github.com/hybridlabor-api/bdb-agent-orchestrator)
|
|
639
|
-
[](LICENSE)
|
|
299
|
+
## MCP servers
|
|
640
300
|
|
|
641
|
-
|
|
301
|
+
[`mcp_config.json`](mcp_config.json) defines <!-- count:mcps -->21<!-- /count --> servers, built or warmed by the installer from `mcps/` and merged into each harness's MCP configuration. Each server exposes tools for a specific domain; every harness sees the same set, avoiding per-tool incompatibilities.
|
|
642
302
|
|
|
643
|
-
|
|
644
|
-
|
|
303
|
+
**Creative software integrations** (primary and fallback pairs for redundancy):
|
|
304
|
+
- **Unreal Engine** — `bdb_unreal_mcp` (Web Remote Control API on port 30010), skill: `bdb-unreal-mcp`
|
|
305
|
+
- **Rhino 3D & Grasshopper** — `bdb_rhino_mcp` (McNeel's Yak router) + `bdb_rhino_mcp_fallback` (GOLEM 3D, 105 tools), skill: `bdb-rhino-mcp`
|
|
306
|
+
- **DaVinci Resolve** — `bdb_davinci_mcp` (workspace scripts, 162 tools) + `bdb_davinci_mcp_studio` (Node.js for Studio) + `bdb_davinci_mcp_fallback`, skill: `bdb-davinci-mcp`
|
|
307
|
+
- **Blender** — `bdb_blender_mcp` (socket integration) + `bdb_blender_mcp_fallback`, skill: `bdb-blender-mcp`
|
|
308
|
+
- **After Effects** — `bdb_after_effects_mcp` + `bdb_after_effects_mcp_fallback`, skill: `bdb-after-effects-mcp`
|
|
309
|
+
- **TouchDesigner** — `bdb_touchdesigner_mcp` (MindDesigner bridge on port 9980) + `bdb_touchdesigner_mcp_fallback`, skill: `bdb-touchdesigner-mcp`
|
|
310
|
+
- **Additional:** grandMA3 (OSC/UDP on port 8000), Resolume (REST API on port 8080), Vectorworks (semantic RAG on port 8765), Adobe UXP bridge, Open Design
|
|
645
311
|
|
|
646
|
-
|
|
647
|
-
|
|
648
|
-
|
|
649
|
-
B --> C[Agent Session 1: Feature Build]
|
|
650
|
-
B --> D[Agent Session 2: Refactoring]
|
|
651
|
-
B --> E[Agent Session N: Test & Verification]
|
|
652
|
-
C --> F[Live Terminal Control & Process Monitor]
|
|
653
|
-
D --> F
|
|
654
|
-
E --> F
|
|
655
|
-
F --> G[Automatic CI/CD Feedback Loops]
|
|
656
|
-
G --> H[PR Review & Merge Routing]
|
|
657
|
-
H --> I[Central Git Repository]
|
|
658
|
-
```
|
|
312
|
+
**OS control & system automation:**
|
|
313
|
+
- **macOS/Linux** — `zavora_computer_use` (native Rust NAPI binary, no runtime compile), skill: `bdb-computer-use-mcp`
|
|
314
|
+
- **Windows** — `bdb_windows_computer_use` (Win32 / COM / UIAutomation, local OCR with Tesseract)
|
|
659
315
|
|
|
660
|
-
|
|
661
|
-
|
|
316
|
+
**Memory, delegation & infrastructure:**
|
|
317
|
+
- **memB** — `memb_mcp` (local offline vector memory, SQLite + ONNX model)
|
|
318
|
+
- **deja** — local transcript indexing (secrets redacted)
|
|
319
|
+
- **mcsc** — multi-harness task delegation
|
|
320
|
+
- **GitHub** — native MCP tools for issues, PRs, workflows
|
|
321
|
+
- **Chrome DevTools** — browser automation & debugging
|
|
322
|
+
- **RemoteOS** — multi-cloud execution gateway with 4-eyes approval engine
|
|
662
323
|
|
|
663
|
-
|
|
664
|
-
- **Desktop Meta-Harness:** Coordinates multi-workspace setups, environment variables, and local server ports across concurrent developer environments.
|
|
665
|
-
- **Parallel Agent Execution:** Spawns autonomous agents working simultaneously on separate modules, features, or bug fixes without interfering with the primary workspace branch.
|
|
666
|
-
</details>
|
|
324
|
+
---
|
|
667
325
|
|
|
668
|
-
|
|
669
|
-
<summary><strong>🔬 Technical Specifications & Automated Routing</strong></summary>
|
|
326
|
+
## Optional modules
|
|
670
327
|
|
|
671
|
-
|
|
672
|
-
- **Automatic CI/CD Feedback Loops:** Monitors test outputs and build tasks, routing error traces directly back into the executing subagent's context for instant repair.
|
|
673
|
-
- **PR Review Routing:** Packages completed features, runs automated security and code health checks, and routes generated Pull Requests for user review or automated merging.
|
|
674
|
-
</details>
|
|
328
|
+
The installer's module picker offers, and Quick Update keeps current. All are optional; AOS works standalone without any of them.
|
|
675
329
|
|
|
676
|
-
|
|
677
|
-
<summary><strong>🔌 Platform Support, Supported Harnesses & Direct Repository Link</strong></summary>
|
|
330
|
+
### memB — Local Vector Memory
|
|
678
331
|
|
|
679
|
-
-
|
|
680
|
-
- **Supported Agent Harnesses:**
|
|
681
|
-
- **Google Antigravity / AGY CLI**
|
|
682
|
-
- **Claude Desktop & Claude Code**
|
|
683
|
-
- **Cursor & Windsurf**
|
|
684
|
-
- **Roo Code & Cline**
|
|
685
|
-
- **ChatGPT Codex / Codex CLI**
|
|
686
|
-
- **Aider & VS Code**
|
|
687
|
-
- **Direct Repository:** Access the orchestrator at [github.com/hybridlabor-api/bdb-agent-orchestrator](https://github.com/hybridlabor-api/bdb-agent-orchestrator).
|
|
332
|
+
`@hybridlabor-api/memb`: offline, local vector memory with an MCP server, WebUI on port 8088, and an ambient hook that injects relevant memories into Claude Code sessions. Skill: `memb-skill`, `memb-ingest`, `bdb-memb-mcp`.
|
|
688
333
|
|
|
689
|
-
|
|
690
|
-
git clone https://github.com/hybridlabor-api/bdb-agent-orchestrator.git
|
|
691
|
-
```
|
|
692
|
-
</details>
|
|
334
|
+
### deja — Transcript Indexing
|
|
693
335
|
|
|
694
|
-
|
|
336
|
+
`@vshulcz/deja-vu`: indexes your agent transcripts locally (secrets redacted), with `deja fix` on an error, `deja wip` to resume, `deja search` for past sessions. Installed with memB. Skill: `deja-memory`.
|
|
695
337
|
|
|
696
|
-
|
|
338
|
+
### OpenWiki — Living Documentation
|
|
697
339
|
|
|
698
|
-
[
|
|
699
|
-
[](https://github.com/hybridlabor-api/bdb-synapse)
|
|
700
|
-
[](https://github.com/hybridlabor-api/bdb-synapse)
|
|
701
|
-
[](https://github.com/hybridlabor-api/bdb-synapse/blob/main/LICENSE)
|
|
340
|
+
`openwiki` CLI: generates and refreshes a grounded wiki of a codebase, with a visualizer on port 4321 and a background daemon. Skill: `openwiki-skill`. This repo's wiki: [.openwiki/](.openwiki/quickstart.md).
|
|
702
341
|
|
|
703
|
-
|
|
342
|
+
### Synapse — 3D Code City
|
|
704
343
|
|
|
705
|
-
|
|
344
|
+
`@hybridlabor-api/bdb-synapse`: renders a repository as a 3D code city and replays agent sessions as light trails. Skill: `synapse-integration-skill`.
|
|
706
345
|
|
|
707
346
|
```mermaid
|
|
708
347
|
flowchart LR
|
|
@@ -714,39 +353,27 @@ flowchart LR
|
|
|
714
353
|
G --> H[Interactive 3D Code City]
|
|
715
354
|
```
|
|
716
355
|
|
|
717
|
-
###
|
|
718
|
-
|
|
719
|
-
| Agent | Log Source | Status |
|
|
720
|
-
|-------|-----------|--------|
|
|
721
|
-
| **Claude Code** | `~/.claude/projects/` | ✅ Native |
|
|
722
|
-
| **Codex CLI** | `~/.codex/sessions/` | ✅ Native |
|
|
723
|
-
| **Pi Agent** | `~/.pi/agent/sessions/` | ✅ Native |
|
|
724
|
-
| **Antigravity (agy)** | `~/.gemini/antigravity-cli/brain/` | ✅ BDB Extension |
|
|
356
|
+
### AO — Agent Orchestrator
|
|
725
357
|
|
|
726
|
-
|
|
727
|
-
- **Tree & Terrain Views:** Repository as a radial tree or treemap — glow ∝ how deeply a file was touched.
|
|
728
|
-
- **Touch States:** Seen (moss green), Read (moonlight blue), Edited (warm amber), Unvisited (dark).
|
|
729
|
-
- **Playback Deck:** Scrub or play the session over a bucketed histogram. Observation stays cool, mutation glows warm.
|
|
730
|
-
- **Agent Lenses:** When a session launched subagents, pick a lens to replay any subagent's trace on the same map.
|
|
731
|
-
- **Session Evaluation:** Ask a local agent CLI to judge the session's trajectory against criteria drafted from your own request.
|
|
358
|
+
`@hybridlabor-api/bdb-agent-orchestrator`: parallel agents in Git worktrees with live terminal control and automated CI/CD feedback loops. Skill: `agent-orchestrator`.
|
|
732
359
|
|
|
733
|
-
```
|
|
734
|
-
|
|
735
|
-
|
|
736
|
-
|
|
360
|
+
```mermaid
|
|
361
|
+
flowchart TD
|
|
362
|
+
A[Desktop IDE Meta-Harness] --> B[Git Worktree Orchestrator]
|
|
363
|
+
B --> C[Agent Session 1: Feature Build]
|
|
364
|
+
B --> D[Agent Session 2: Refactoring]
|
|
365
|
+
B --> E[Agent Session N: Test & Verification]
|
|
366
|
+
C --> F[Live Terminal Control & Process Monitor]
|
|
367
|
+
D --> F
|
|
368
|
+
E --> F
|
|
369
|
+
F --> G[Automatic CI/CD Feedback Loops]
|
|
370
|
+
G --> H[PR Review & Merge Routing]
|
|
371
|
+
H --> I[Central Git Repository]
|
|
737
372
|
```
|
|
738
373
|
|
|
739
|
-
|
|
740
|
-
|
|
741
|
-
## 🎨 BDB Creator Extension: Heavy-Lifting Media & 3D Compute Pipeline
|
|
742
|
-
|
|
743
|
-
[](https://github.com/hybridlabor-api/bdb-dev-creator-extension)
|
|
744
|
-
[](https://github.com/hybridlabor-api/bdb-dev-creator-extension)
|
|
745
|
-
[](https://github.com/hybridlabor-api/bdb-dev-creator-extension)
|
|
746
|
-
[](https://github.com/hybridlabor-api/bdb-dev-creator-extension)
|
|
747
|
-
[](LICENSE)
|
|
374
|
+
### Creator Extension — Media & 3D
|
|
748
375
|
|
|
749
|
-
|
|
376
|
+
`@hybridlabor-api/bdb-dev-creator-extension`: ComfyUI MCP capabilities (FLUX, SDXL), image-to-3D (TripoSR, TRELLIS), and automated video production (OpenMontage, Remotion). Skill: `bdb-dev-creator-extension`.
|
|
750
377
|
|
|
751
378
|
```mermaid
|
|
752
379
|
flowchart LR
|
|
@@ -755,76 +382,18 @@ flowchart LR
|
|
|
755
382
|
B --> D[Cinema Video Suite]
|
|
756
383
|
B --> E[Local ComfyUI MCP Engine]
|
|
757
384
|
C --> C1[TRELLIS: High-Fidelity 3D]
|
|
758
|
-
C --> C2[TripoSR: Fast Mesh
|
|
385
|
+
C --> C2[TripoSR: Fast Mesh]
|
|
759
386
|
C --> C3[CadQuery: Text-to-CAD]
|
|
760
387
|
D --> D1[OpenMontage AI Director]
|
|
761
388
|
D --> D2[Remotion Video-Shotcraft]
|
|
762
|
-
D --> D3[Palmier Pro NLE MCP Server]
|
|
763
389
|
E --> E1[FLUX.1 Image Gen]
|
|
764
390
|
E --> E2[SDXL Pipeline]
|
|
765
|
-
|
|
766
|
-
C1 & C2 & C3 & D1 & D2 & D3 & E1 & E2 & E3 --> F[Rendered Media & Spatial Assets]
|
|
391
|
+
C1 & C2 & C3 & D1 & D2 & E1 & E2 --> F[Rendered Media & Spatial Assets]
|
|
767
392
|
```
|
|
768
393
|
|
|
769
|
-
|
|
770
|
-
<summary><strong>🔷 3D Generation Suite (`engines/3d/`)</strong></summary>
|
|
771
|
-
|
|
772
|
-
- **Microsoft TRELLIS:** High-fidelity image-to-3D asset generation producing textured 3D meshes and NeRF/Gaussian Splat representations.
|
|
773
|
-
- **Stability AI TripoSR:** Ultra-fast sub-second (<0.5s) single-image to 3D mesh generation for rapid spatial prototyping.
|
|
774
|
-
- **CadQuery Parametric Text-to-CAD:** Generates precise engineering models and architectural geometry in STEP, STL, and URDF formats.
|
|
775
|
-
</details>
|
|
776
|
-
|
|
777
|
-
<details>
|
|
778
|
-
<summary><strong>🎬 Cinema Video Suite (`engines/video/`)</strong></summary>
|
|
779
|
-
|
|
780
|
-
- **OpenMontage AI Orchestrator:** Automated video storytelling, script-to-timeline assembly, and shot sequencing.
|
|
781
|
-
- **Remotion Video-Shotcraft:** 100+ cinema-grade programmatic video components and motion graphics templates built with React.
|
|
782
|
-
- **Palmier Pro NLE MCP Server:** Real-time HTTP MCP bridge (`http://127.0.0.1:19789/mcp`) exposing macOS native non-linear video editing capabilities.
|
|
783
|
-
</details>
|
|
784
|
-
|
|
785
|
-
<details>
|
|
786
|
-
<summary><strong>🎨 Local ComfyUI MCP Engine & Direct Repository Link (`mcps/comfyui-mcp/`)</strong></summary>
|
|
787
|
-
|
|
788
|
-
- **Model Context Protocol Integration:** Exposes local ComfyUI workflows directly as executable tools for AI agents over MCP.
|
|
789
|
-
- **Supported Generative Models:**
|
|
790
|
-
- **FLUX.1:** High-resolution image synthesis and prompt adherence.
|
|
791
|
-
- **SDXL:** Latent diffusion workflow control with custom LoRAs and ControlNets.
|
|
792
|
-
- **Wan2.1:** Generative video diffusion models for high-frame-rate clip creation.
|
|
793
|
-
- **Direct Repository:** Access the extension suite at [github.com/hybridlabor-api/bdb-dev-creator-extension](https://github.com/hybridlabor-api/bdb-dev-creator-extension).
|
|
794
|
-
|
|
795
|
-
```bash
|
|
796
|
-
git clone https://github.com/hybridlabor-api/bdb-dev-creator-extension.git
|
|
797
|
-
```
|
|
798
|
-
</details>
|
|
799
|
-
|
|
800
|
-
<details>
|
|
801
|
-
<summary><strong>🌍 BDB OS Remote Gateway & Thin-Client (`mcps/bdb-os-remote/`)</strong></summary>
|
|
802
|
-
|
|
803
|
-
- **Zero-Trust SSE Transport:** Run Claude Desktop on your laptop while executing tools natively on your stationary Workstation over a secure Tailscale tunnel.
|
|
804
|
-
- **Asymmetric Topology:** Installs `heimdall-token-saver` locally on your laptop to compress tokens *before* calling the LLM, while `memB`, `synapse`, and file operations are routed to the Workstation.
|
|
805
|
-
- **Offline Clone Tool:** One-click project archives streamed seamlessly over Tailscale without heavy `node_modules`.
|
|
806
|
-
- **Direct Repository:** Access the gateway at [github.com/hybridlabor-api/bdb-os-remote](https://github.com/hybridlabor-api/bdb-os-remote) or install via NPX:
|
|
807
|
-
|
|
808
|
-
```bash
|
|
809
|
-
npx @hybridlabor-api/bdb-os-remote installer
|
|
810
|
-
```
|
|
811
|
-
</details>
|
|
812
|
-
|
|
813
|
-
---
|
|
814
|
-
|
|
815
|
-
## ⚡ BDB Hardware & PCB: Electrical & Enclosure Design Module
|
|
816
|
-
|
|
817
|
-
[](https://github.com/hybridlabor-api/bdb-hardware-pcb)
|
|
818
|
-
[](https://github.com/hybridlabor-api/bdb-hardware-pcb)
|
|
819
|
-
[](https://www.kicad.org/)
|
|
820
|
-
[](LICENSE)
|
|
394
|
+
### Hardware & PCB — Electrical Design
|
|
821
395
|
|
|
822
|
-
|
|
823
|
-
schematic capture, PCB layout/routing, DFM sign-off, and parametric 3D
|
|
824
|
-
enclosure design into the agent loop — governed by the core
|
|
825
|
-
[`godmode-hardware-pcb`](#-the-7-godmodes-apex-layer) persona, which owns the
|
|
826
|
-
IPC-standard trace/impedance math and the headless KiCad ERC/DRC/DFM gate no
|
|
827
|
-
board is allowed to skip on its way to fabrication.
|
|
396
|
+
`@hybridlabor-api/bdb-hardware-pcb`: KiCad and OpenSCAD design module, driven by `godmode-hardware-pcb` skill. Exposes ERC/DRC gate, gerber sign-off, and parametric enclosure design. Skills: `godmode-hardware-pcb`, `bdb-hardware-pcb`.
|
|
828
397
|
|
|
829
398
|
```mermaid
|
|
830
399
|
flowchart LR
|
|
@@ -833,162 +402,56 @@ flowchart LR
|
|
|
833
402
|
B --> D[Schematic Capture & ERC]
|
|
834
403
|
B --> E[PCB Layout & Routing]
|
|
835
404
|
B --> F[DRC / DFM / Gerber Sign-Off]
|
|
836
|
-
C --> G[Parametric Enclosure
|
|
405
|
+
C --> G[Parametric Enclosure]
|
|
837
406
|
D & E & F & G --> H[Fabrication-Ready Output]
|
|
838
407
|
```
|
|
839
408
|
|
|
840
|
-
###
|
|
841
|
-
- **`kicad-mcp-server`** — 47 tools over stdio JSON-RPC for schematic capture, ERC, PCB layout, routing, DRC, and Gerber/BOM/CPL export. Targets **KiCad 9 & 10** via `kicad-cli`.
|
|
842
|
-
- **`openscad-mcp-server`** — 8 tools for parametric 3D enclosure and mechanical co-design (OpenSCAD + BOSL2).
|
|
843
|
-
|
|
844
|
-
### 5 Skills (`category: engineering-hardware`)
|
|
845
|
-
|
|
846
|
-
| Skill | Description |
|
|
847
|
-
|-------|-------------|
|
|
848
|
-
| `schematic-datasheet-analysis` | Electrical rule auditing, datasheet grounding, pinout validation, power tree tracing, and negative-evidence analysis for KiCad schematics. |
|
|
849
|
-
| `pcb-constraint-definition` | Translates high-level hardware requirements into formal engineering constraints, layer stackup, netclasses, and custom DRC rules for KiCad. |
|
|
850
|
-
| `pcb-layout-routing-automation` | Floorplanning, placement rules, high-speed differential pair routing, return-path continuity, thermal via arrays, and keepout enforcement. |
|
|
851
|
-
| `pcb-validation-dfm-signoff` | Automated DRC/ERC verification, SI/PI screening, fab-house DFM/DFA compliance, and production release sign-off. |
|
|
852
|
-
| `code-first-hardware-design` | Programmatic schematic capture and circuit synthesis (SKiDL, text netlists, S-expressions) plus parametric 3D enclosure co-design. |
|
|
853
|
-
|
|
854
|
-
### Install
|
|
855
|
-
|
|
856
|
-
Same optional-module mechanism as `bdb-synapse` and
|
|
857
|
-
`bdb-dev-creator-extension` — the main AOS installer downloads and runs it for
|
|
858
|
-
you, or install it standalone on macOS, Linux, or Windows:
|
|
859
|
-
|
|
860
|
-
```bash
|
|
861
|
-
npx @hybridlabor-api/bdb-hardware-pcb
|
|
862
|
-
```
|
|
863
|
-
|
|
864
|
-
Supported harnesses: **Claude Code**, **OpenAI Codex**, and **Google
|
|
865
|
-
Antigravity (Gemini)**. Verify a local install with:
|
|
866
|
-
|
|
867
|
-
```bash
|
|
868
|
-
./scripts/test_mcp_connection.sh --all
|
|
869
|
-
```
|
|
870
|
-
|
|
871
|
-
---
|
|
872
|
-
|
|
873
|
-
## 🧠 memB: Custom Semantic Brain
|
|
874
|
-
|
|
875
|
-
BDB OS introduces a fully integrated local, offline-first semantic memory brain based on **memB**. It provides zero-compute context for SLMs and an AI-first flat-file vault architecture.
|
|
876
|
-
|
|
877
|
-
<details>
|
|
878
|
-
<summary><strong>⚙️ How the Ecosystem Works (Skills, Vaults, & Obsidian)</strong></summary>
|
|
879
|
-
|
|
880
|
-
### 1. Ingestion via the `/memb-ingest` Skill
|
|
881
|
-
The ecosystem includes a deeply integrated skill (`/memb-ingest`). When an agent runs this, the `memb_ingest.py` script recursively scans your project (reading `.openwiki`, `AGENTS.md`, transcripts, and architecture files).
|
|
882
|
-
* **Offline Vector Embeddings:** It bundles a pre-quantized 30MB `all-MiniLM-L6-v2` ONNX model to chunk and store these learnings natively in a fast SQLite vector store (`~/.MemBDB/memb.db`), all without hitting external APIs.
|
|
883
|
-
|
|
884
|
-
### 2. Autonomous AI-First Vault Generation
|
|
885
|
-
Once ingestion completes, memB natively generates a **physical Markdown Vault** (`~/.MemBDB/memB_Vault`) structured around a strict "God Mode" radial topology:
|
|
886
|
-
* **Zero-Compute Context:** A universal `AGENTS.md` and a master `God_Mode.md` are written to the root. Small 30MB local inference models can instantly orient themselves macroscopically by reading these physical files without spending context tokens on complex database calls.
|
|
887
|
-
* **Micro-Targeted RAG:** For precise execution, the small LLMs query the vector DB to retrieve just the exact 3-5 sub-files needed.
|
|
888
|
-
|
|
889
|
-
### 3. The Obsidian Visualization Plugin
|
|
890
|
-
memB includes a native **Obsidian Plugin** (`obsidian-memb-plugin`) that acts as a visual UI over your physical vault.
|
|
891
|
-
* **Top-Down Radial Tree:** It reads the generated `memB_Vault` and maps it visually using Obsidian's graph view.
|
|
892
|
-
* By strictly using directional parent-to-child links (e.g., God Mode -> Projects -> Category -> Neuron), the Obsidian graph blossoms outward like a flower, completely preventing the "black hole" context clustering seen in unstructured graph databases.
|
|
893
|
-
|
|
894
|
-
### 4. Data Sovereignty & Security
|
|
895
|
-
* **Zero Telemetry:** Absolute data sovereignty with no remote tracking.
|
|
896
|
-
* **Secret Filtration:** Blocks passwords, raw API keys, and connection strings before injection.
|
|
897
|
-
</details>
|
|
409
|
+
### Heimdall Token Saver — CLI Output Compression
|
|
898
410
|
|
|
899
|
-
|
|
900
|
-
|
|
901
|
-
## ⚡ Heimdall Token Saver: CLI Context Compression
|
|
902
|
-
|
|
903
|
-
|
|
904
|
-
|
|
905
|
-
**Heimdall Token Saver** is an ultra-fast context compression engine designed to drastically reduce context window usage for CLI tool execution outputs in AI agent workflows.
|
|
906
|
-
|
|
907
|
-
<details>
|
|
908
|
-
<summary><strong>⚙️ Purpose & Performance</strong></summary>
|
|
909
|
-
|
|
910
|
-
- **Automatic CLI Output Context Compression:** Reduces token overhead by **60–99%** on high-volume CLI tool outputs without impacting agent understanding.
|
|
911
|
-
- **Zero Information Loss Guarantee:** Preserves all error messages, failed assertions, stack traces, exit codes, and actionable debugging context while stripping redundant whitespace, progress spinners, and repetitive logs.
|
|
912
|
-
- **Automatic Secret Redaction:** Automatically detects and redacts passwords, tokens, API keys, and sensitive environment variables prior to inserting command output into agent context windows.
|
|
913
|
-
</details>
|
|
411
|
+
`@hybridlabor-api/heimdall-token-saver`: compresses repeated CLI output via ambient hooks on every harness. Reduces token overhead on large projects. Skill: `token-saver-config`.
|
|
914
412
|
|
|
915
|
-
|
|
916
|
-
<summary><strong>🔬 Technical Specifications & Processors</strong></summary>
|
|
413
|
+
|
|
917
414
|
|
|
918
|
-
|
|
919
|
-
- **Version Control & Dev Tools:** `git` (status, diff, log, branch)
|
|
920
|
-
- **Testing Frameworks:** `pytest`, `jest`, `cargo test`, `vitest`, `go test`
|
|
921
|
-
- **Containers & Infrastructure:** `docker`, `kubectl`, `terraform`
|
|
922
|
-
- **Package Managers & Build Systems:** `npm`, `yarn`, `pnpm`, `pip`, `cargo`, `go` package listings and build outputs
|
|
923
|
-
- **Preservation Rules:** Guarantees line numbers, error traces, and exact failure sites remain 100% intact for immediate root-cause diagnosis.
|
|
924
|
-
</details>
|
|
925
|
-
|
|
926
|
-
<details>
|
|
927
|
-
<summary><strong>🔌 Agent Integration & Hooks</strong></summary>
|
|
928
|
-
|
|
929
|
-
- **Automated Hook Installation:** Configured and installed seamlessly via `installer.js`.
|
|
930
|
-
- **Supported Harnesses:**
|
|
931
|
-
- **Claude Code:** Integrated via `PreToolUse` hook.
|
|
932
|
-
- **Google Antigravity CLI:** Integrated via `AfterTool` hook.
|
|
933
|
-
|
|
934
|
-
### 📊 CLI Diagnostics & Tooling
|
|
935
|
-
You can run diagnostic and benchmarking commands directly in your terminal:
|
|
936
|
-
- **Check Version:** `token-saver version`
|
|
937
|
-
- **View Savings & Usage Statistics:** `token-saver stats`
|
|
938
|
-
- **Benchmark Command Savings:** `token-saver benchmark '<command>'`
|
|
939
|
-
</details>
|
|
415
|
+
*Illustrative sketch from v3.x. The percentages are the project's own estimates, not measurements made for this README.*
|
|
940
416
|
|
|
941
417
|
---
|
|
942
418
|
|
|
943
|
-
##
|
|
944
|
-
|
|
945
|
-
### 🆚 Which Version Should I Use?
|
|
946
|
-
|
|
947
|
-
- **`@latest`** — the stable channel. Interactive MCP selection UI, active background daemons (`memB`, `OpenWiki`), full skill library.
|
|
948
|
-
- **`@next`** — an ad-hoc staging channel used occasionally for large changes (like the v4.0.0 AOS rename) before promotion to `@latest` (which CI publishes to automatically on every release-please release). It is not a permanent parallel channel.
|
|
949
|
-
- **`bdb-antigravity-skills@legacy`** — the original, pre-dispatcher Antigravity-only pack, kept for anyone still depending on it.
|
|
419
|
+
## Updating
|
|
950
420
|
|
|
951
|
-
|
|
952
|
-
|
|
953
|
-
### Ask Your AI Agent (Easiest)
|
|
954
|
-
Tell your assistant: *"Run `npx -y @hybridlabor-api/aos@latest` to install the skills pack and configure the local MCP servers."*
|
|
955
|
-
|
|
956
|
-
### Command Line
|
|
421
|
+
Run the same command again. The installer sees the installed version, offers **Quick Update**, and refreshes skills, hooks, templates and modules:
|
|
957
422
|
|
|
958
423
|
```bash
|
|
959
|
-
# Stable
|
|
960
424
|
npx -y @hybridlabor-api/aos@latest
|
|
961
|
-
|
|
962
|
-
# Migration staging channel (v4.0.0 AOS)
|
|
963
|
-
# Note: @next is only used occasionally for active migrations, not a standing channel.
|
|
964
|
-
npx -y @hybridlabor-api/aos@next
|
|
965
425
|
```
|
|
966
|
-
*(Works on Mac/Linux terminals as well as Windows PowerShell.)*
|
|
967
426
|
|
|
968
|
-
|
|
427
|
+
There is no `aos update` subcommand. If you once ran `npm i -g @hybridlabor-api/aos`, a plain `aos` on your PATH runs that frozen copy and its version, not the latest; either update it (`npm i -g @hybridlabor-api/aos@latest`) or remove it and stay with `npx`. The installer prints the update command itself whenever a newer version exists; the `bdb-updater` skill wraps the same check for use from inside a session.
|
|
969
428
|
|
|
970
|
-
|
|
429
|
+
## Uninstall
|
|
971
430
|
|
|
972
431
|
```bash
|
|
973
|
-
|
|
432
|
+
aos-uninstall # removes what AOS installed; memory, wikis and credentials stay
|
|
433
|
+
aos-uninstall --purge # also removes ~/.MemBDB, ~/.openwiki, ~/.synapse, ~/.memb
|
|
434
|
+
aos-uninstall --dry-run # list everything, delete nothing
|
|
974
435
|
```
|
|
975
436
|
|
|
976
|
-
|
|
437
|
+
The uninstaller works from the install manifest: a file that still matches the hash AOS wrote is removed, a file you edited is backed up instead, a file AOS never wrote is not touched. The same action is in the installer menu.
|
|
977
438
|
|
|
978
|
-
|
|
439
|
+
---
|
|
979
440
|
|
|
980
|
-
|
|
981
|
-
npx -y @hybridlabor-api/aos --project-harness
|
|
982
|
-
```
|
|
441
|
+
## Contributing
|
|
983
442
|
|
|
984
|
-
|
|
443
|
+
- [AGENTS.md](AGENTS.md) is the single source of rules for every harness: the skill contract, category routing, the release gate, Conventional Commits.
|
|
444
|
+
- A skill is a directory with `SKILL.md`; the frontmatter needs `name` (equal to the directory), `description` and `category`. `npm run validate` enforces the contract, as CI does on every push.
|
|
445
|
+
- `npm test` runs the validator self-test, the plugin-manifest check and the installer, store, doctor and cross-harness hook tests.
|
|
446
|
+
- `.claude-plugin/plugin.json` and `marketplace.json` are generated by `npm run plugin:build` and checked by `npm run plugin:check`. They exist today; the Claude Code marketplace install path is still being finalised, so the installer above remains the supported route.
|
|
447
|
+
- Releases are cut by release-please from Conventional Commits; do not bump `package.json` by hand. `feat:` means a minor bump.
|
|
448
|
+
- Skills derived from other projects record `source:` in the frontmatter and an entry in [THIRD_PARTY_NOTICES.md](THIRD_PARTY_NOTICES.md).
|
|
985
449
|
|
|
986
|
-
|
|
987
|
-
git clone https://github.com/hybridlabor-api/aos.git
|
|
988
|
-
cd aos
|
|
989
|
-
npm install
|
|
990
|
-
node installer.js
|
|
991
|
-
```
|
|
450
|
+
## Links
|
|
992
451
|
|
|
993
|
-
|
|
994
|
-
|
|
452
|
+
- Package: [npmjs.com/package/@hybridlabor-api/aos](https://www.npmjs.com/package/@hybridlabor-api/aos)
|
|
453
|
+
- Source and issues: [github.com/hybridlabor-api/aos](https://github.com/hybridlabor-api/aos) · [issues](https://github.com/hybridlabor-api/aos/issues)
|
|
454
|
+
- [CHANGELOG.md](CHANGELOG.md) · [docs/skills_table.md](docs/skills_table.md)
|
|
455
|
+
- Sibling repos: [bdb-agent-orchestrator](https://github.com/hybridlabor-api/bdb-agent-orchestrator) · [bdb-synapse](https://github.com/hybridlabor-api/bdb-synapse) · [bdb-dev-creator-extension](https://github.com/hybridlabor-api/bdb-dev-creator-extension) · [bdb-hardware-pcb](https://github.com/hybridlabor-api/bdb-hardware-pcb) · [bdb-os-remote](https://github.com/hybridlabor-api/bdb-os-remote)
|
|
456
|
+
|
|
457
|
+
License: [Apache-2.0](LICENSE).
|