Pipeline API

Programmatic access to the Live Log Insight pipeline.

Pipeline API

Programmatic access to run Live Log Insight pipelines.

Python API

Basic Usage

from src.agentX.pipeline import run_pipeline

# Run full pipeline
result = run_pipeline(
    time_range="24h",
    environment="production",
    service_name="auth-service"
)

# Access results
print(result.summary)
print(result.anomalies)
print(result.recommendations)

Advanced Usage

from src.agentX.pipeline import Pipeline

# Create custom pipeline
pipeline = Pipeline(
    config_path="config.json",
    output_dir="output"
)

# Run specific stages
pipeline.run_stages(["discover", "fetch", "parse", "aggregate"])

# Access stage outputs
metrics = pipeline.get_output("aggregate")

CLI API

Command Line Interface

# Full pipeline
uv run main.py --time-range 24h --environment production

# With options
uv run main.py \
  --time-range 12h \
  --environment staging \
  --service-name payment-service \
  --severity-filter ERROR \
  --max-logs 5000

Output API

Accessing Outputs

from src.agentX.pipeline import PipelineOutput

# Load outputs
outputs = PipelineOutput.load("output/")

# Access individual outputs
summary = outputs.summary
anomalies = outputs.anomalies
metrics = outputs.metrics
hypotheses = outputs.hypotheses
recommendations = outputs.recommendations