Workflows¶

Multi-agent YAML workflows — no Python required.

Discovery (`workflows/discovery/`)¶

File	What it does
`biomodels_discovery_pipeline.yaml`	Question → ranked shortlist
`biomodels_assumption_review.yaml`	Assumption review + human approval
`biomodels_baseline_simulation.yaml`	Baseline simulation
`biomodels_perturbation_study.yaml`	Perturbation scenarios
`biomodels_comparison_report.yaml`	Comparison report
`biomodels_full_research_workflow.yaml`	All phases (7 steps)

praisonai workflow run workflows/discovery/biomodels_discovery_pipeline.yaml

File	Demo
`glycolysis_demo.yaml`	BIOMD0000000206 end-to-end
`mapk_p53_discovery.yaml`	MAPK / p53 shortlist

File	Pattern
`sysbio_route_by_input.yaml`	Route by domain
`parallel_sensitivity_scan.yaml`	Parallel scans
`team_model_review.yaml`	AgentTeam review
`full_platform_showcase.yaml`	Route + Parallel + Memory
`precision_medicine_loop.yaml`	Loop: baseline → perturb → compare

Production patterns — policy, approval, eval, MCP, bots.

File	What it does
`approval_assumption_gate.yaml`	Human approval before assumption review
`approval_parameter_scan.yaml`	Human approval before parameter scan
`policy_guarded_simulation.yaml`	Simulation with export path policy
`eval_regression_gate.yaml`	Eval gate on tool selection
`mcp_sysbio_server.yaml`	Run sysbio MCP server workflow
`botos_sysbio_telegram.yaml`	Telegram bot + sysbio tools (needs token)
`botos_stub.yaml`	BotOS wiring stub

praisonai workflow run workflows/platform/policy_guarded_simulation.yaml

Agent defaults: workflows/agents/defaults.yaml

graph TD
    A[Intake] --> B[Search BioModels]
    B --> C[Assumption review]
    C --> D[Simulate]
    D --> E[Report]

Before running: python -c "import praisonai_bio"