level-progression

name	level-progression
description	THE HYPER BRAIN level tracker — 1–20 progression, what each level represents, what's unlocked, what's pending (Level 18–20). Use when the user says "level X", "level up", "what's next", "level 18", "constellation", or asks about Brain progression status.

level-progression

The Brain is a 20-level system. Each level = a feature/capability shipped. Levels 1–17 done as of May 7, 2026. Levels 18–20 in flight.

Level Tracker (current)

✅ 1–8   Vault scaffold + plugins + PARA structure
✅ 9     GitHub bridge (4hr polling) — scripts/github_to_obsidian.py
✅ 10    Vault immortal (Obsidian Git auto-commit, 10min)
✅ 11    BROski$ Coin Tracker (Dataview widget)
✅ 12    Hyperfocus CSS Modes (Focus / Calm / Hyper)
✅ 13    Morning Briefing AI                  ✅ May 7
✅ 14    GitHub Webhooks real-time            ✅ May 7
✅ 15    HyperAgent MCP Bridge                ✅ May 7
✅ 16    Focus Tracker + Analytics            ✅ May 7
✅ 17    HyperSplit Task Decomposition        ✅ May 7
⏳ 18    AI Distraction Filter (wired to focus sessions)   ← NEXT
⏳ 19    DifficultyDial + Dynamic XP
⏳ 20    THE HYPER BRAIN Constellation (MCP mesh + RAG)

What Each Level Unlocks

Foundational (1–8)

The vault structure, Obsidian plugins, PARA folders. Done in early sprints. vault-para-structure skill covers this.

Level 9 — GitHub Bridge

scripts/github_to_obsidian.py polls GitHub every 4 hours, syncs issues + PRs into 00-Inbox/GitHub/<repo>/. Cron-driven, not real-time.

Level 10 — Vault Immortal

Obsidian Git plugin auto-commits the vault every 10 mins, auto-pushes to remote. Means the vault never loses a note even if a machine dies.

Level 11 — BROski$ Coin Tracker

Dataview query renders a coin balance widget in the Dashboard note. Reads from vault frontmatter (coins: field).

Level 12 — Hyperfocus CSS Modes

Three CSS themes:

• focus-mode — minimal, distraction-free
• calm-mode — softer colors, longer sessions
• hyper-mode — high contrast, fast scanning

Toggle via Obsidian command palette.

Level 13 — Morning Briefing AI ✅ May 7

/briefing/generate → AI summarizes overnight GitHub activity, today's tasks, focus stats → drops into 00-Inbox/Briefings/<date>.md. Runs on demand or via cron.

Level 14 — GitHub Webhooks Real-Time ✅ May 7

/webhook/github receives GitHub events live. New issue → instant note in 00-Inbox/GitHub/<repo>/. Replaces 4hr polling for connected repos.

Level 15 — HyperAgent MCP Bridge ✅ May 7

/mcp/query exposes the vault as an MCP source. HyperAgent agents (V2.4 stack) can query vault notes via MCP gateway port 8820.

Level 16 — Focus Tracker + Analytics ✅ May 7

/focus/start + /focus/end log sessions to 05-Focus-Sessions/. /analytics/weekly + /analytics/heatmap produce reports.

Level 17 — HyperSplit Task Decomposition ✅ May 7

/hypersplit recursively breaks a task into micro-tasks (default depth 3). Output → 01-Projects/<project>/HyperSplit-<task>.md with linked sub-tasks.

Level 18 — AI Distraction Filter (NEXT)

Code DONE: ai_distraction_filter.py exists, has /distraction/report + /distraction/patterns. Needs wiring into focus sessions — when a session is active, distractions logged via the filter feed into the session's record.

The work:

1. In focus_tracker.py /focus/start: register session ID + intensity
2. When /distraction/report is called during a session: increment session's distraction count
3. When /focus/end: include distraction summary in the session log
4. Cross-correlate in /analytics/weekly (focus quality score)

Level 19 — DifficultyDial + Dynamic XP

XP per task currently flat. Level 19 = task XP scales with difficulty (HyperSplit's micro-task count, time estimate, distractions). DifficultyDial = a UI knob in the Dashboard to set today's difficulty target (which auto-tunes briefing tone).

Not yet started.

Level 20 — THE HYPER BRAIN Constellation

The endgame: full MCP mesh (Brain ↔ V2.4 ↔ Course ↔ Pets ↔ SDK) with RAG over the vault. Means any agent in the ecosystem can query the vault and receive contextual answers grounded in Bro's notes.

Not yet started. Depends on Level 19.

Confirming Level Status

curl http://localhost:8100/health
# → {"status":"hyper","level":20,"containers":30}
# (level=20 here means "all 8 modules loaded" — not the progression level)

The module-level health (from /health) is different from the progression level. Health=20 means engine green. Progression level is hand-tracked in CLAUDE_CONTEXT.md.

Update The Tracker

When a level ships:

1. Update CLAUDE_CONTEXT.md (root) — change ⏳ to ✅, add date
2. Update CLAUDE.md (root) — same
3. Update this skill's table
4. If the level adds endpoints → update hyper-brain-modules skill's module map
5. Commit + let Obsidian Git push automatically

What To Work On Next

Level 18 first — the code is already there, just wiring required. Estimate: 1–2hrs.

After 18:

• Level 19 — design pass first (what defines "difficulty"? what does the dial UI look like?), then implement
• Level 20 — dependent on stable MCP mesh; talk to V2.4's mcp-gateway skill

Companion Skills

• hyper-brain-modules — module-level reference
• morning-briefing-ai — Level 13 deep dive
• vault-para-structure — Levels 1–8 foundation
• obsidian-git-vault — Level 10 mechanics

Hard Rules

• One level at a time — finish 18 before starting 19
• Track levels in CLAUDE_CONTEXT.md — it's the source of truth, this skill mirrors it
• Honest tracker — code "done but not wired" is ⏳, not ✅
• Docker compose path is docker-compose.hyper-brain.yml at root, not docker/