From smart-dispatch
Automatically routes tasks to the optimal AI agent, model, or provider based on complexity, cost, and capability. Use when implementing features, fixing bugs, or any multi-step development work. Triggers on "implement", "build", "create", "fix", "add feature", "develop", or when the user asks to do any coding task.
How this skill is triggered — by the user, by Claude, or both
Slash command
/smart-dispatch:smart-dispatchThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
If an `EXEC-MAP v1` block from `senior-prompt-engineer` is in context, **consume it instead of re-deriving complexity:**
If an EXEC-MAP v1 block from senior-prompt-engineer is in context, consume it instead of re-deriving complexity:
EXEC-MAP.effort seeds the tier (trivial/low → Tier 0; medium/high → Tier 1).EXEC-MAP.models is the per-phase routing intent (plan→Opus/quality, impl→Sonnet/balanced, mechanical→Haiku/budget); honor it unless validation proves it wrong.EXEC-MAP.executor picks the provider/CLI; map model tiers to that CLI's tiers.EXEC-MAP.mcp lists tools to wire up.
Treat the map as a starting routing decision, not gospel — escalation rules (Tier 1) still override it when validation keeps failing. With no EXEC-MAP, derive routing from the request as usual.Before dispatching to an AI agent for Tier 0 tasks, attempt to solve it using local CLI tools.
npm run lint --fix or eslint --fix.prettier --write.tsc and check if errors are trivial.If local tools fail or aren't enough, use Gemini-Flash YOLO via RTK.
rtk gemini --yolo.Any agent dispatched in YOLO/Bypass mode MUST execute the relevant validation command before completion.
tsc --noEmit.npm run lint.npm run test <file>.Always pick the cheapest tier that passes validation; escalate per 1.1. Tiers map per provider:
| Tier | Claude | Codex | Gemini |
|---|---|---|---|
| budget | Haiku 4.5 | gpt low/mini reasoning | Flash |
| balanced | Sonnet 4.6 | gpt medium reasoning | Pro |
| quality | Opus 4.8 / Fable 5 | gpt high reasoning | Pro (max thinking) |
| Task type | Tier | Notes |
|---|---|---|
| Docs, comments, changelogs, summaries, translations | budget | Haiku-class is enough; never burn Opus here |
| Test writing — mechanical / known pattern | budget | Verify with test run (0.3 mandate) |
| Lint / typecheck / format fixes | local CLI → budget | Tier 0 rules apply first |
| Implementation, refactor, multi-step agentic | balanced | Sonnet-class default for code |
| Test design — integration/e2e strategy | balanced | Strategy needs reasoning; writing the cases can drop to budget |
| Plan, architecture, root-cause, security audit, migration | quality | Opus-class; deep reasoning pays for itself |
EXEC-MAP's models: {plan, impl, mechanical} maps 1:1 to quality/balanced/budget rows above.
If a Tier 0 agent fails to resolve the issue (validation still fails) after 2 attempts, the dispatcher MUST:
| Domain | Best Platform | Reason |
|---|---|---|
| Multimodal / UI screenshots | Gemini Pro (latest) | Best image/screenshot interpretation. |
| Complex Logic / Math | Codex (gpt high reasoning) | Superior reasoning for algorithms. |
| Agentic Chains / Auth | Claude Sonnet 4.6+ | Best tool-use and multi-step autonomy. |
| Deep Plan / Architecture | Claude Opus 4.8 / Fable 5 | Strongest long-horizon reasoning. |
| Open Source / Docs | OpenCode | Deep integration with community patterns. |
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub andersonlimahw/lemon-ai-hub --plugin smart-dispatch