Back to corpus
technical notebacklog referencescore 18

Compass Implementation Roadmap

1. **Mermaid Architecture Diagram** — Already done (architecture.mmd) 2. **State Collection Script** — `collect-state.sh` gathering all system state to JSON 3. **Pipeline Dependency Map** — Document all 23 pipeline dependencies 4. **Skill Category Index** — Parse all 136 skill SKILL.md files into a searchable index 5. **Service Health Script** — Single script that checks all 4 core services

Open full HTMLDownload HTMLCorpus JSONPaper schema

Full HTML reader

Read the full artifact

Open in new tab

Extracted abstract or opening context

**Goal:** A Next.js dashboard that reads from existing system state and displays everything in one place. ### Week 1: Foundation - [ ] Initialize Next.js 15 project with App Router, Tailwind, shadcn/ui - [ ] Create `collect-state.sh` script that gathers: - `clawdbot cron list` output (pipeline state) - Service health checks (curl Graph Kernel, RAG++, Clawdbot Gateway) - `git status` for all Desktop projects - Skill file listing from `[home-path]` - Registry.yaml parse - [ ] Build data layer (`lib/data-sources.ts`) that reads collected state - [ ] **Pipeline Dashboard** — Table view of all 23 cron jobs with status, last run, next run - [ ] **Service Status** — Health cards for Graph Kernel, RAG++, Orbit, Clawdbot Gateway ### Week 2: Visualization - [ ] **Architecture Graph** — D3.js + Dagre visualization of project categories and connections - [ ] **Skill Registry** — Browse all 136 skills by category with search - [ ] **Project Health** — Git status, last commit, CI status for Desktop projects - [ ] Auto-refresh via polling (30-second interval) - [ ] Dark mode (default) with system preference detection - [ ] Deploy locally with `npm run dev` — accessible on Tailscale **Goal:** Proper pipeline orchestration with dependencies, retries, and monitoring. ### Week 3: Dagster Setup - [ ] Install Dagster locally (`pip install dagster dagster-webserver`) - [ ] Define Dagster project structure in `compass/dagster/` - [ ] Create Python wrappers for 5 low-priority pipelines: - Garden Tender - HTDS Rebuild - Daily MFP Post - Research Digest - Daydream Scanner - [ ] Configure retry policies per pipeline - [ ] Run in shadow mode alongside existing crons - [ ] Verify Dagit UI is accessible

Promotion decision

What has to happen next

Keep in the searchable backlog until it intersects a live paper or system.

Why this is not always a full paper yet

Corpus pages are public-safe readers for discovered workspace artifacts. They are not automatically final papers. A corpus item becomes a polished paper only after the editable source, evidence checkpoints, references, figures, render path, and release status are attached through the paper schema.