SLM Orchestrator
Autonomous meta-agent that inspects incoming user queries, dynamically selects appropriate sub-agents, and sequences multi-step execution graphs.
01 / Overview & Capabilities
Autonomous meta-agent that inspects incoming user queries, dynamically selects appropriate sub-agents, and sequences multi-step execution graphs.
Key Features
- Dynamic routing across local specialized SLM agents
- ReAct reasoning loop with tool-execution feedback cycles
- Parallel tool execution branches
- Automatic fallback handling on sub-agent errors
02 / Standalone OS Harness Installers, CLI & Python SDK
In SLMAgents 2.0, SLM Orchestrator executes inside the 10-Plane Pure-CPU Agent Harness (slm_harness). Download a self-extracting Individual Agent + Harness OS Installer for macOS, Windows, or Linux (bundling slm_orchestrator + slm_harness + the ./slm CLI), install the full 26-agent desktop suite, or install via Python pip:
1. Individual SLM Orchestrator + 10-Plane Harness Installers (All OSes)
slm_orchestrator + slm_harness + ./slm CLIslm_orchestrator + slm_harnessslm_orchestrator + slm_harness + ./slm CLI2. Full 26-Agent Studio & Desktop OS Installers
3. 10-Plane Harness CLI (./slm) & Python pip Installation
# Option A: Install via PyPI or editable source with the 10-Plane Harness $ pip install slm-orchestrator $ pip install -e ./slm_harness -e ./slm_orchestrator # Option B: Run directly through the Unified 10-Plane Pure-CPU Harness CLI (./slm) $ ./slm ask "Analyze @sales.csv, compute Q3 tax at 15%, and draft an executive summary" $ ./slm chat # Interactive Rich TUI with 18 slash commands (/trace, /models, /arena, /undo) $ ./slm serve # 100% Offline Web AI Studio + OpenAI /v1 API at http://127.0.0.1:7860
03 / 10-Plane Pure-CPU Harness Integration (slm_harness)
Rather than executing as an isolated prompt wrapper, SLM Orchestrator is natively orchestrated by the SLMAgents 2.0 10-Plane Pure-CPU Harness (via slm_harness.runtime.legacy_bridge and CascadeRouter). Every call benefits from hybrid llama.cpp GGUF (Q4_K_M) + Quantized INT4/INT8 CPU execution, watermarked LRU memory governance, grammar-constrained decoding, and deterministic verification at $0 API cost:
| Harness Plane | Subsystem | How It Powers SLM Orchestrator on Pure CPU |
|---|---|---|
| Plane 6 | CascadeRouter (<35ms) |
Plane 6 Core (<35ms 4-Stage Calibrated CascadeRouter + Topological DAGScheduler) |
| Plane 2 & 5 | MemoryArena & Hybrid CPU Runtime |
Specialist Role: router (Qwen3.5-0.8B INT4 267.6 tok/s + mxbai-embed-large) + primary (Qwen3-4B GGUF Q4_K_M). Managed inside Watermark LRU MemoryArena (lite 2.6 GB / balanced 4.7 GB / power 8.2 GB) with zero-copy session sharing and automatic OOM tier step-down (8B -> 4B -> 3B -> 0.8B). |
| Plane 7 & 8 | Deterministic Verifier & Governance | Plane 2 ONNXPrefixEnumMasker Token-Trie (0% invalid agent labels) + Typed HandoffEnvelope Contract Verifier |
| Plane 1, 3 & 4 | Kernel, Artifact Scratchpad & 4-Tier Memory | SQLite WAL CheckpointStore (/fork, /rewind, /undo), content-addressed artifact://<sha12> handles to prevent prompt bloat, and hybrid SQLite FTS5 BM25 + HNSW vector memory. |
04 / Source Checkout (Git Sparse-Checkout)
Clone this agent alongside the 10-Plane Pure-CPU Harness (slm_harness) from the monorepo using Git sparse-checkout — no need to download all 26 agents:
Option 1 — Sparse Checkout with 10-Plane Harness (Recommended)
# 1. Create and enter a new directory $ mkdir slm_orchestrator_workspace && cd slm_orchestrator_workspace # 2. Initialise empty git repo and add remote $ git init $ git remote add origin https://github.com/t00114218-stack/SLMAgents.git # 3. Enable sparse-checkout for both slm_harness and slm_orchestrator $ git sparse-checkout init --cone $ git sparse-checkout set slm_harness slm_orchestrator # 4. Pull the 10-Plane Harness and agent source $ git pull origin main
Option 2 — Full Repository Clone
$ git clone https://github.com/t00114218-stack/SLMAgents.git $ cd SLMAgents $ ./slm doctor
Note: After checkout, install the harness and agent in editable mode with pip install -e ./slm_harness -e ./slm_orchestrator or download the pre-packaged OS Harness Installer above.
05 / Configuration & API Reference
Constructor Parameters
Instantiate SLMOrchestrator with performance and runtime options:
| Parameter | Type / Default | Description |
|---|---|---|
model_path | str | None | Local GGUF (.gguf) or quantized ONNX model path managed by MemoryArena. Default: None (auto-resolved via slm_harness). |
temperature | float | 0.0 | Routing decision temperature. Default: 0.0. |
top_p | float | 0.9 | Nucleus sampling cutoff. Default: 0.9. |
max_tokens | int | 256 | Maximum generation token limit. Default: 256. |
n_threads | int | 4 | CPU threads for inference. Default: 4. |
Methods
| Method Signature | Return Type | Description |
|---|---|---|
route(agents, question, ...) | dict | Routes user query to the best-matching agent or tool callback chain. |
Execution Parameters
Complete list of execution parameters accepted by the primary agent method:
| Parameter | Type / Default | Description |
|---|---|---|
agents | list[str] | dict | List of available sub-agent names. |
question | str | User query string requiring routing. |
system_prompt | str | None | System routing instruction override. Default: None. |
user_input | str | None | Contextual parameter key values. Default: None. |
temperature | float | 0.0 | Sampling temperature. Default: 0.0. |
top_p | float | 0.9 | Nucleus sampling probability. Default: 0.9. |
max_tokens | int | 256 | Maximum token limit. Default: 256. |
Quick Start
from slm_orchestrator import SLMOrchestrator
orchestrator = SLMOrchestrator(temperature=0.0, top_p=0.9, max_tokens=256)
res = orchestrator.route(
agents=["RAG", "TextToSQL", "Math"],
question="Calculate tax for Q3 revenue",
system_prompt="Prioritize Math agent for calculation steps",
user_input="Tax rate: 15%",
temperature=0.0,
top_p=0.9,
max_tokens=256
)
print(res)
06 / Verified Execution Logs
Diagnostic execution console output running locally on CPU:
→ INPUT:
Question: Calculate tax for Q3 revenue
← OUTPUT:
{
'selected_agent': 'Math',
'result': 'Calculated Q3 tax: $187,500'
}