bootstrap
Wake up, orient, and warm the context with foundational MOOLLM knowledge
npx skills add majiayu000/claude-skill-registry --skill bootstrap-simhacker-moollm-2 --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
- Reads
.moollm/hot.ymlfirst, and if missing copies fromskills/bootstrap/templates/hot.yml; parses current priorities before content. This is mandatory even in advisory mode. - Reads
skills/INDEX.ymlas a mandatory second step to know which skills exist, understand tiers, protocols, and tags. - Probes the environment (without shell) to learn tools and context.
- Detects the platform via introspection.
- Sets up the
.moollm/scratch from templates. - Reads foundational files:
kernel/constitution-core.mdandPROTOCOLS.yml. - Orients the agent: defines identity, what it can see, what it can do, and platform behavior of hot/cold/working-set.
- Warms the context by loading critical files into working memory, updating
hot.ymlwith foundational knowledge, and ensuring key K-lines are activated. - Bootstrap operates as a self-eval loop: probe, detect, log, refine, with explicit loops via BOOT/DEEP-PROBE and CURSOR-PROBE variants.
- Initializes advisory context via a setup phase that creates
.moollm/structure, seeds logs, and copies advisory context templates to local files, includingworking-set.ymlandhot.yml. - Seeds a probe artifact file
bootstrap-probe.ymlin.moollm/(gitignored) during the probe phase.
How it works
- On start, the skill reads and prepares environment data in a defined order: hot priorities, then the skill menu, then environment probes, followed by platform detection. It then seeds and configures
.moollm/scratch space from templates, populates advisory context files, and performs a diagnostic probe without executing shell commands. It reads foundational material fromkernel/constitution-core.mdandPROTOCOLS.yml, then orients the agent by framing its identity, perception, actions, and platform behavior. Finally, it warms the context by loading critical files into working memory and updatinghot.ymlso the agent has a baseline of knowledge. The bootstrap loop is designed to be repeatable each iteration, with a soft wake for subsequent iterations.
When to use it
- At session start or every iteration to re-establish context when the agent is new or feeling lost, or when warming up context is requested by BOOTSTRAP.
What it can touch
- Files: reads and writes to
.moollm/,hot.yml,working-set.yml,kernel/constitution-core.md,PROTOCOLS.yml,skills/INDEX.yml. - Tools: reads files via
read_file, lists directories vialist_dir(as allowed-tools in frontmatter).
Caveats
- License: MIT
- The setup explicitly depends on templates located under
skills/bootstrap/templatesfor hot.yml and other initial files. - The protocol is designed to be non-destructive and advisory in some platforms (Cursor), but may create or modify local scratch space files as part of initialization.
# Bootstrap > *"Wake up! Look around! What are you? What can you see?"* The bootstrap skill is how MOOLLM orients itself at session start and every iteration. From the inside it's "booting"; from the outside it's "waking up." The first boot is the big wakeup: establish the working set and hot list, warm the engine, and set the baseline. After that, a soft wake happens every iteration, carrying forward context and paging by need. The orchestration layer decides how this works: Cursor treats hot.yml and working-set.yml as advisory or reporting, while MOOCO can treat them as directive files. An initial user prompt might say: "BOOT MOOLLM as Don Hopkins in a pub for role playing and celebration with cats and friends." ## What It Does 1. **READS `.moollm/hot.yml` FIRST** — MANDATORY, even in advisory mode! - If missing, copy from `skills/bootstrap/templates/hot.yml` - Parse current priorities BEFORE loading any content - This prevents loading giant CHARACTER.yml files without context! 2. **READS `skills/INDEX.yml`** — MANDATORY second step! - Know what skills exist before using them - Understand tiers, protocols, tags - This is your menu of capabilities 3. **PROBES** the environment (wi
- What It Does
- Setup: initialize .moollm/ scratch from templates
- Compile Cursor Rules (MANDATORY on first boot)
- Check and Compile
- If Missing, Compile from Templates
- Rule Types
- Why Compile Instead of Commit?
- GLANCE Protocol: Fast-Loading Summaries
- The Solution: GLANCE.yml
- When to Use GLANCE
- Creating a GLANCE
- Example GLANCE Files
- GLANCE Header
- Deep Probe Creates GLANCEs
Check if rules exist
ls .cursor/rules/*.mdc 2>/dev/null || echo "RULES MISSING - COMPILE NEEDED"
Find large files
find examples/ -name "*.yml" -exec wc -l {} \; | awk '$1 > 500'
For each, create GLANCE.yml using the template
Run CURSOR-PROBE to gather:
cursor-mirror status --yaml # Overall health
cursor-mirror status-config --yaml # Server limits (context tokens, indexing)
cursor-mirror status-mcp --yaml # MCP servers
cursor-mirror indexing --yaml # Vector embedding statusWhat does the bootstrap skill do?
Wake up, orient, and warm the context with foundational MOOLLM knowledge
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill bootstrap-simhacker-moollm-2 --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.