Catlx Runtime Lifecycle

CATLX — Runtime Lifecycle Sequences

This skill owns the ordered end-to-end sequences that tie the subsystems together: how CATLX boots, how a voice command flows, how a workflow runs, how a plugin is installed, and how recovery proceeds.

Canonical detail: ../../knowledge/references/runtime-lifecycle.md. The voice sequence is also in ../../examples/voice-command-lifecycle.md. Load on demand.

Purpose

Give the exact, ordered step sequences so that no stage is skipped or performed out of order. Use these as the authoritative procedural truth for orchestrating CATLX behavior.

When to activate

• User asks about startup/boot order, or "what happens when CATLX starts."
• Tracing a voice command, a workflow run, a plugin install, or a recovery end-to-end.
• Verifying a specific stage (e.g. when the Capability Router runs vs the Memory Broker).

What this skill handles

1. Boot sequence (§18.1) — launcher detects CATLX_ROOT → stale-PID check → Hardware Profiler (< 50 ms) → Capability Router emits CapabilityMap → databases open (WAL replay) → Module Registry scanned in dependency order → Event Bus on :7701 → Security Layer → Memory Broker → AI Provider Layer (health/local servers T1+) → Workflow Engine scheduler → Voice Pipeline (wake word + mic) → Plugin Runtime → Electron Shell/HUD → write PID lockfile → emit system.ready → restore workspace → operational.
2. Voice command lifecycle (§18.2) — wake word → capture → VAD (800 ms) → STT → context → NLU (IntentObject, confidence < 0.75 → disambiguate/loop) → Workflow Planner → Permission validation → DAG execute → per-step checkpoint + TTS/HUD → completion → Memory Broker writes EpisodicRecord → monitoring.
3. Workflow lifecycle (§18.3) — load YAML/dynamic → validate DAG (no cycles, deps present, modules available) → create execution context (run ID, input snapshot, rollback log) → schedule → dispatch root nodes → per-node module invoke → store output/unlock dependents → on failure retry policy → on FAILED offer rollback → on success COMPLETED → emit completion → archive run.
4. Plugin lifecycle (§18.4) — package (marketplace/local) → signature verify → manifest parse → capability grant review → store grants in plugins.db → resolve/vendor deps → build/validate → register → load into sandbox → initialize() → register capabilities with router → available → update (hot-swap or restart flag) → uninstall (cleanup(), deregister, destroy sandbox, remove files).
5. Recovery lifecycle (§18.5) — crash detected via stale PID → Recovery Mode flag / Safe Mode UI → replay WAL → load workspace snapshot → find RUNNING workflows → verify checkpoint hash → present Resume/Rollback/ Ignore → execute chosen action → clear flag → continue boot.

Requirements / constraints

• R8 (replay safety): all lifecycle steps that mutate state are checkpointed.
• R7 (deterministic recovery): recovery sequence is deterministic; Recovery Mode flag cleared only when complete.
• Windows-first paths and mechanisms apply throughout (see ../../knowledge/rules/windows-rules.md).

Canonical knowledge it reads

../../knowledge/references/runtime-lifecycle.md · ../../knowledge/references/recovery.md · ../../knowledge/references/workflow-engine.md · ../../knowledge/references/voice-pipeline.md · ../../knowledge/references/plugin-ecosystem.md · ../../knowledge/references/data-registries.md.

Delegation

Because these sequences cross subsystems, delegate to the owning skill when a specific stage needs detail: catlx-voice-pipeline, catlx-workflow-engine, catlx-plugin-ecosystem, catlx-electron-shell, catlx-security, catlx-recovery, catlx-hardware-adaptation, catlx-capability-routing. Use skill({ name: "<target>" }) and pass the specific stage as the minimum context.

Edge cases & warnings

• Boot ordering: the Capability Router must emit the CapabilityMap before any subsystem initializes.
• Stale PID: a genuine crash triggers recovery; a live PID must not be treated as stale.
• Confidence loop: NLU must loop on disambiguation only while confidence < threshold; avoid infinite loops (cap the number of clarification turns).
• Plugin/unit ordering at boot: modules and plugins load in dependency order.

Component lifecycle policy (reuse → install → adapt → create)

NEVER create a new component as the default. Before building/creating anything (a sub-skill, dependency, reference, workflow, helper, adapter, or template), check, in order:

1. Reuse an existing local component (resolve aliases/equivalent capabilities first) — reuse, don't rebuild.
2. Use an already-registered component from the registry.
3. Install a suitable existing, trusted, supported component → validate → register → connect to the graph → use.
4. Adapt an existing compatible component via a small persistent adapter/wrapper instead of re-creating it.
5. Create only as last resort — then make it permanent immediately: stable id, canonical location, register, add to the capability index + dependency graph, add provenance, use, and allow future reuse.
6. Never reorganise/recreate already-generated components (no Skill X 2 / new / temp variants); extend the existing one. Never create a second competing knowledge source; connect back to the canonical knowledge/ layer. Promote any reusable artifact out of /tmp/scratch into the permanent ecosystem.

Full policy: ../../knowledge/rules/component-lifecycle.md.

Source / provenance

• Source: PART XVIII §18.1–18.5 (boot, voice command lifecycle detailed, workflow lifecycle, plugin lifecycle, recovery lifecycle).
• Inferred: none; sequences preserved as ordered steps.