Scbe Internet Workflow Synthesis

Build deterministic internet workflow pipelines from existing SCBE assets and the AI-Workflow-Architect repo, then run post-baseline variable tuning.

Quick Start

1. Synthesize a repo-local profile:

python C:/Users/issda/.codex/skills/scbe-internet-workflow-synthesis/scripts/synthesize_pipeline_profile.py \
  --repo-root C:/Users/issda/SCBE-AETHERMOORE \
  --output training/internet_workflow_profile.json \
  --force

2. Run baseline plus variable tuning:

python C:/Users/issda/.codex/skills/scbe-internet-workflow-synthesis/scripts/run_e2e_pipeline.py \
  --repo-root C:/Users/issda/SCBE-AETHERMOORE \
  --profile training/internet_workflow_profile.json

Workflow

1. Read references/workflow-template-map.md and prioritize local templates first.
2. Generate or update training/internet_workflow_profile.json.
3. Execute baseline with scripts/web_research_training_pipeline.py.
4. Read the latest summary.json emitted by the baseline run.
5. Tune thresholds and runtime knobs using scripts/tune_system_variables.py.
6. Emit a tuned cloud-kernel config plus a tuning report.

Output Contract

Return or persist these artifacts:

• Baseline profile JSON (training/internet_workflow_profile.json)
• Baseline run summary (training/runs/web_research/<run_id>/summary.json)
• Tuned thresholds config (training/cloud_kernel_pipeline.tuned.json by default)
• Tuning report (artifacts/internet_workflow_tuning_report.json by default)
• Tuned runtime profile (training/internet_workflow_profile.tuned.json by default)

Invariants

• Keep governance checks enabled for production flows (skip_core_check=false).
• Keep deterministic audit artifacts (summary.json, audit.json, decision_record.json).
• Keep secrets in environment variables; do not write tokens into profile files.
• Keep template discovery explicit (local first, GitHub fallback).

Resource Guide

• scripts/synthesize_pipeline_profile.py: Build a default profile from template sources.
• scripts/run_e2e_pipeline.py: Run baseline ingest then invoke tuning automatically.
• scripts/tune_system_variables.py: Tune thresholds and runtime knobs from baseline metrics.
• references/workflow-template-map.md: Canonical local and GitHub template file map.
• references/system-variable-tuning.md: Tuning policy and variable bounds.
• assets/internet_workflow_profile.template.json: Copyable starter profile.