From next-task
Classifies tasks by cognitive tier (sm0l/ch0nky/frontier) and routes to optimal model from b00t config, preferring local/cheap for mechanical work and frontier for reasoning. Checks resources.
npx claudepluginhub elasticdotventures/_b00t_ --plugin next-taskThis skill uses the workspace's default tool permissions.
NEVER use frontier model for mechanical work. Route by cognitive tier.
Routes coding tasks to optimal AI model tier by complexity: no LLM for mechanical edits, Haiku for simple refactors, Sonnet for multi-file bugs, Opus for architecture/security. Saves 50-65% API costs.
Recommends Claude models (Haiku for exploration, Sonnet for implementation, Opus for decisions) via routing matrix for task types, subagents, and cost-quality tradeoffs.
Routes OpenRouter API calls to optimal models by task (e.g., code review to Claude-3.5-Sonnet) or prompt complexity for cost, quality, latency optimization in multi-model apps.
Share bugs, ideas, or general feedback.
NEVER use frontier model for mechanical work. Route by cognitive tier.
Reads routing table from _b00t_/model-routing.tomllm. Falls back to hardcoded tiers.
| Tier | Tasks | b00t Models (in priority order) |
|---|---|---|
small (sm0l) | grep, lint, classify, route, test pass/fail | haiku, local sm0l |
chunky (ch0nky) | implement, refactor, debug, code review | qwen3-coder-local (RTX 3090), sonnet |
frontier | architecture, security, novel design, planning | opus, sonnet |
b00t learn model-routing via MCP or CLI — NEVER read .tomllm directly. # output: available_models[]b00t hive status — ensure RAM/GPU available. # output: resource_ok{model, tier, rationale} for caller to invoke.small sm0l (Haiku / local 3B):
chunky ch0nky (qwen3-coder-local → Sonnet fallback):
frontier (Opus → Sonnet fallback):
Executive context is costly. Sub-agents MUST return compressed summaries:
| Tier | Max output to executive |
|---|---|
sm0l | PASS or FAIL: <name> <5 lines> |
ch0nky | diff + test result (no full file dumps) |
frontier | structured decision with rationale |
Before invoking ch0nky/frontier check hive:
b00t hive status # output: RAM free, GPU VRAM free, active profile
Anti-pattern: running vLLM (qwen3-coder, 20GB VRAM) + HuggingFace download simultaneously on 24GB.
Used by /next-task at each phase to select model.
Used by b00t-mcp agent delegation.
Load via: b00t learn model-routing (MCP preferred, CLI fallback)