Skip to content

Manifest reference

MAS Lab is configured through YAML manifests: declarative files for agents, multi-agent systems, experiments, datasets, overlays, and pipelines.

New to the vocabulary? Read glossary.md first (scenario, overlay, pipeline, run, flavour).

Hands-on: start with Tutorial 1 (agent) → Tutorial 2 (MAS) → Tutorial 3 (experiments).

How manifests fit together

experiment.yaml          ← what to run (scenarios × dataset × n_runs + pipelines)
    │
    ├── applications[]   → mas.yaml / registered app
    ├── scenarios[]    → overlay stacks per variant
    ├── dataset        → prompts, turns, memory seeds
    └── application.post / scenario.post / …  → pipeline steps (metrics, plots)

mas.yaml                 ← team topology, workflow, transport
    └── agents/*.yaml    ← design pattern, tools, skills, observability

config.yaml       ← project defaults (flavour, infra_refs, .env path)
Layer Manifest kinds Reference
Agent Agent agent.md
MAS MAS, Workflow mas.md, workflow.md
Override Overlay overlay.md
Environment Flavour, InfraBundle, LLMProxy flavour.md, infra.md
Experiment experiment: experiment.md
Inputs Dataset dataset.md
Processing pipeline: / kind: Pipeline pipeline.md
Interactive demo lab: lab.md

Runtime execution manifests (Agent, MAS, overlays, infra) are documented under runtime.md.

Resolution: ref vs id

Form Example Meaning
File path ref: ./agents/broker.yaml Relative to the referring manifest
Catalog name dataset.name: arborian-network Resolved under the lab or library
Library ref standard:openai Bundled infra from library-standard
CLI override --infra-ref, -o overlay.yaml One-shot for mas-ctl

Schema files

YAML schemas live under docs/schemas/. The bench UI and mas-ctl validate compile them at runtime.

Area Schema directory
Lab (experiment, dataset, pipeline) docs/schemas/lab/
Runtime (agent, mas, overlay, infra) docs/schemas/runtime/
Workspace config.schema.yaml

Application binding (tutorials vs paper labs)

Both forms are valid in experiment.applications[]:

Style Example When to use
Inline manifest manifest: ./agent.yaml + optional configs_dir Tutorials, self-contained experiments
Registered app app: trip-planner + configs_dir: overlays/ Paper labs, shared apps under library-samples/apps/

Scenarios reference overlay ids from configs_dir (e.g. tutorial cot vs lab pattern-cot). Dataset: path: ./dataset.yaml (tutorial) or name + locator: samples (catalogued benchmarks).

See topology-and-workflow.md for workflow vs routing overlays.

See also