Bumblebiber

Update flow for hmem. Runs `npm update -g`, syncs skills, applies migrations, verifies hooks, shows the changelog. Use when the user asks to update/upgrade hmem, hmem-mcp, or its-over-9k (any language), or when the startup version-check flags a new release. Runs the npm update itself — don't assume it's already done.

Switch active project mid-session via load_project, then fix any misrouted O-entry exchanges. Use when the user says 'aktiviere/lade Projekt X', 'switch to project Y', or whenever you'd otherwise call load_project mid-session.

View and change hmem memory settings, hooks, sync, and checkpoints. Use on /hmem-config, or when the user asks why context is huge, how to change auto-save, how often checkpoints fire, or how to set up sync.

Merge all session checkpoint summaries into one final O-Entry summary. Run at end of sessions with 2 or more checkpoints.

Load specific context from hmem based on what is needed RIGHT NOW. Use when load_project output is not enough for the current question.

Curate an .hmem file (your own or foreign) — mark obsolete/irrelevant, fix titles, consolidate duplicates, repair broken links. **Requires the `hmem-curate` MCP server** (skill prompts the user to enable it on entry). Use whenever the user says 'aufräumen', 'memory aufräumen', 'Speicher aufräumen', 'hmem aufräumen', 'clean up memory', 'tidy up hmem', 'curate memory', 'consolidate duplicates', 'merge duplicate entries', 'fix broken links', 'kümmer dich um die Memory', or invokes /hmem-curate. Also trigger when `memory_health()` flags BLOCKER/WARNING issues, when a P-entry's load_project output exceeds 4k tokens (session-start noise check defers to this skill), or before any batch cleanup of L, E, D, P entries. Skipping this skill and editing memory directly bypasses health checks, severity triage, and obsolete-chain integrity — never curate without it.

Dispatch an isolated sub-agent that returns only [RESULT]...[/RESULT]. Use for any search, lookup, 'does X exist?', 'find Y', calculation, or isolated writing — never do these inline when dispatch fits.

Migrate O-entries to the v5.3+ project-bound 5-level structure (each O00XX bound to its P00XX). Use on 'migrate O-entries', when hmem-update flags a structure migration, or when you notice O-entries still in the old flat L2→L4→L5 format.

Guided first-install and reconfiguration flow for o9k. Use when the user invokes /o9k-init, on a fresh o9k install, or when the user wants to set up, reconfigure, or extend their companion stack. Detects what's installed, interviews the user (bundle, git, conflicts), resolves collisions with migration, then executes the setup.

Token usage report for the current project from Claude Code session transcripts. Use when the user asks how many tokens were used/saved, invokes /o9k-stats, or wants to verify the o9k pillars are paying off.

Meta-skill for the o9k doctrine and arbitration — who owns output style, the repo map, dispatch, memory, the plan. Load at session start, on conflicts between efficiency skills, and whenever unsure Path A vs Path B — if a team-up roster exists, external CLI tmux workers require team-up + dispatch mailbox/watcher (not bare dispatch). Also use when caveman vs other output-styles collide.

Personalized o9k orientation for the human. Use when the user invokes /o9k-guide, asks how o9k works or what they need to do, right after first-time setup, or when the SessionStart hook flags an unresolved arbitration. Explains what runs automatically, the few one-time actions left, and on-demand commands (/o9k-init, /o9k-stats, /o9k-update). Agent-only: if a team-up roster exists, external CLI spawns still require dispatch path B — this guide does not exempt that.

Check installed o9k pillars and companion frameworks for updates, and apply the safe ones. Use when the user invokes /o9k-update, asks whether anything is out of date, or acts on the SessionStart 'updates available' notice.

Cost-gated subagent dispatch for context isolation. Use for broad searches, lookups, log analysis, doc digestion, independent subtasks, OR whenever you spawn an external CLI worker via the team-up companion (planner/implementer/reviewer in tmux, cursor-agent/claude/codex/hermes). Path A = in-host RESULT subagents. Path B = mailbox run + team-up dispatch --run-id + cheap in-host watcher (runs wait) — REQUIRED for every external CLI tmux spawn when a team-up roster exists; bare dispatch without a watcher is an incomplete spawn (parent never gets notified).

Install a conflict-free stack of third-party o9k companion frameworks in one command. MUST use when the user asks to set up companions, install the recommended stack, add memory/docs/methodology/symbol tools alongside o9k, or extend an existing bundle — never hand-install MCP servers without checking arbitration first. Wraps install/o9k-companions.sh and enforces the one-owner-per-concern rule from using-o9k.

GitHub Scout for agent frameworks. MUST run before installing an unknown companion, extending a bundle, refreshing the o9k compatibility matrix, or answering 'what's new/better than X' for Claude Code plugins, MCP servers, or agent frameworks. Repeatable recon loop: hunt, score, classify against using-o9k arbitration, trial in sandbox, report — never dump raw search results into context.