Python SDK

pip install frootai

from frootai import FrootAIClient
 
client = FrootAIClient()
 
# Search FROOT knowledge base
results = client.search_knowledge("RAG chunking strategies")
for r in results:
    print(f"{r['module']}: {r['title']}")
 
# Get a specific module
module = client.get_module("R2")
print(module["title"])  # "RAG Architecture"
 
# List all solution plays
plays = client.list_plays()
print(f"{len(plays)} plays available")

# Search across all 17 FROOT modules
results = client.search_knowledge(query: str, max_results: int = 5)
 
# Get full content of a specific module
module = client.get_module(module_id: str)
# module_id: F1, F2, F3, R1, R2, R3, O1-O6, T1-T3
 
# List all modules organized by FROOT layer
modules = client.list_modules()
 
# Look up an AI/ML term in the glossary
definition = client.lookup_term(term: str)

# List all solution plays
plays = client.list_plays(filter: str = None)
 
# Get detailed play information
play = client.get_play_detail(play_number: str)
 
# Semantic search for plays
matches = client.search_plays(query: str, top_k: int = 3)
 
# Compare plays side-by-side
comparison = client.compare_plays(plays: list[str])

# Compare AI models for a use case
comparison = client.compare_models(use_case: str, priority: str = "quality")
 
# Get model catalog
catalog = client.get_model_catalog(category: str = "all")
 
# Estimate monthly Azure costs
estimate = client.estimate_cost(play: str, scale: str = "dev")

# Builder agent — implementation guidelines
guidelines = client.agent_build(task: str)
 
# Reviewer agent — security + quality checklist
review = client.agent_review(context: str = None)
 
# Tuner agent — production readiness validation
validation = client.agent_tune(context: str = None)
 
# Run evaluation against thresholds
result = client.run_evaluation(scores: dict, thresholds: dict = None)

from frootai import FrootAIClient
 
client = FrootAIClient()
 
# Run evaluation
scores = {
    "groundedness": 4.5,
    "relevance": 3.8,
    "coherence": 4.1,
    "fluency": 4.6
}
 
result = client.run_evaluation(
    scores=scores,
    thresholds={"groundedness": 4.0, "relevance": 3.5}
)
 
print(f"Overall: {'PASS' if result['overall_pass'] else 'FAIL'}")
for metric in result["results"]:
    status = "✅" if metric["passed"] else "❌"
    print(f"  {status} {metric['metric']}: {metric['score']}")

pip install frootai-mcp
python -m frootai.mcp