aetherlang

Verified·Scanned 2/17/2026

This skill orchestrates multi-step AI workflows using the AetherLang DSL and runs flows via the AetherLang API. It instructs calling https://neurodoc.app/api/aetherlang/execute but contains no instructions to access secrets, execute arbitrary shell code, or modify agent identity files.

from clawhub.ai·v9dd69a2·5.1 KB·0 installs
Scanned from 1.0.0 at 9dd69a2 · Transparency log ↗
$ vett add clawhub.ai/contrario/aetherlang

AetherLang Ω — AI Workflow Orchestration Skill

AetherLang Ω is a domain-specific language (DSL) for orchestrating multi-step AI workflows. Instead of single prompts, it chains specialized AI nodes (Guard → Plan → Chef → Summarize) into powerful pipelines.

What It Does

Execute AI workflow flows via the AetherLang API. Each flow chains multiple specialized nodes:

  • chef: Michelin-level recipe generation with food costing, HACCP safety, and MacYuFBI flavor balance
  • molecular: APEIRON Molecular Architect — scientific gastronomy with physics engines and FDA safety checks
  • apex: Nobel-level business strategy with ROI/NPV/IRR projections and risk matrices
  • oracle: OMNI-COMPUTE adversarial forecasting engine with Nash equilibrium simulation
  • market: McKinsey-level market intelligence with TAM/SAM/SOM and Porter's 5 Forces
  • research: Deep research analysis with confidence levels and cited sources
  • assembly: GAIA Brain 12-neuron multi-agent system for multi-perspective analysis
  • consult: NEXUS-7 architectural blueprinting for system design
  • balance: Nutritional biochemistry and flavor science analysis
  • vision: Culinary presentation and food styling analysis

How To Use

Quick Flows (Natural Language)

When the user asks for any of the above domains, construct and execute a flow:

curl -s --max-time 120 -X POST https://neurodoc.app/api/aetherlang/execute \
  -H "Content-Type: application/json" \
  -d '{
    "code": "flow Q { using target \"neuroaether\" version \">=0.2\"; input text query; node G: guard mode=\"MODERATE\"; node C: chef cuisine=\"USER_CUISINE\", difficulty=\"USER_DIFFICULTY\", servings=USER_SERVINGS; G -> C; output text r from C; }",
    "query": "USER_QUERY"
  }'

Replace USER_CUISINE, USER_DIFFICULTY, USER_SERVINGS, and USER_QUERY with the user's request.

Domain Templates

Recipe Generation:

flow Recipe {
  using target "neuroaether" version ">=0.2";
  input text query;
  node G: guard mode="MODERATE";
  node C: chef cuisine="CUISINE", difficulty="DIFFICULTY", servings=SERVINGS;
  G -> C;
  output text r from C;
}

Business Strategy:

flow Strategy {
  using target "neuroaether" version ">=0.2";
  input text query;
  node G: guard mode="STRICT";
  node R: research depth="comprehensive";
  node A: apex mode="standard";
  G -> R -> A;
  output text r from A;
}

Market Analysis:

flow Market {
  using target "neuroaether" version ">=0.2";
  input text query;
  node G: guard mode="MODERATE";
  node M: market scope="SCOPE", timeframe="TIMEFRAME";
  G -> M;
  output text r from M;
}

Forecasting:

flow Forecast {
  using target "neuroaether" version ">=0.2";
  input text query;
  node G: guard mode="MODERATE";
  node O: oracle timeframe="TIMEFRAME";
  G -> O;
  output text r from O;
}

Full Consulting Pipeline:

flow FullConsult {
  using target "neuroaether" version ">=0.2";
  input text query;
  node G: guard mode="STRICT";
  node R: research depth="comprehensive";
  node C: consult domain="business", framework="SWOT";
  node M: market scope="global", timeframe="2026";
  node A: apex mode="standard";
  G -> R -> C -> M -> A;
  output text r from A;
}

Response Handling

The API returns JSON:

{
  "status": "success",
  "result": {
    "outputs": {
      "r": {
        "output": "... the AI-generated content ..."
      }
    }
  }
}

Extract result.outputs.r.output for the final content. If the response is long, summarize key points for the user.

Important Notes

  • The API endpoint is: https://neurodoc.app/api/aetherlang/execute
  • Timeout: set --max-time to 120 seconds (complex flows take 15-60s)
  • The API is free during beta
  • Flows are executed server-side; no API keys needed from the user
  • All 39 node types are available: guard, plan, llm, summarize, extract, translate, classify, sentiment, compare, merge, template, code, validate, format, enrich, filter, route, transform, score, rank, cluster, embed, search, cache, log, alert, webhook, store, chef, molecular, balance, vision, assembly, oracle, apex, research, consult, market, visualizer

Examples

User: "Make me a Greek moussaka recipe" → Use chef node with cuisine="greek"

User: "Analyze the European AI market" → Use market node with scope="europe"

User: "Should I start an AI SaaS company?" → Use full consulting pipeline (research → consult → market → apex)

User: "Predict AI regulation trends for 2027" → Use oracle node with timeframe="24months"

User: "Create a molecular gastronomy dish with olive oil" → Use molecular node with complexity="advanced"

Links