FORGE Deep Research Pipeline
The FORGE Deep Research (DR) pipeline is a two-phase autonomous system that scans a GitHub organization, produces a structured state audit, and generates an execution plan that runs overnight without human intervention.Pipeline Architecture
Phase 1: Organization State Audit
The State Audit Generator (FORGE-DR-001) scans everything:| Scan Target | What It Checks |
|---|---|
| Repositories | Classification (active, empty, archived, legacy), default branch, visibility |
| Branches | Non-standard defaults, divergence, orphaned feature branches |
| Issues | Open/closed counts, orphans, duplicates, missing labels |
| Pull Requests | Conflicts, stale PRs, duplicate PRs |
| Local Workspace | Clone coverage, branch drift, uncommitted changes |
| Bundles/Epics | Completion validation, hollow completions |
| Cross-Repo | Contract integrity, dangling references, event catalogue consistency |
| Stage | Evidence-backed maturity assessment (PLANNING through PRODUCTION) |
Output Files
- ORGANIZATION_STATE_AUDIT.md — 12-section comprehensive report (15-25 pages)
- state_inventory.yaml — Machine-readable state for Phase 2 consumption
- risk_register.yaml — Prioritized risks with severity, likelihood, mitigation
Phase 2: Autonomous Execution Planner
The Execution Planner (FORGE-EX-001) consumes the audit and produces:Epic-to-Taskset Hierarchy
A2A Handoff Protocol
Every taskset boundary has a formal contract:| Field | Description |
|---|---|
handoff_id | Unique: HO--- |
artifacts | Files/PRs produced by source taskset |
preconditions | What must hold before target starts |
quality_gate | Automated: lint, type-check, test, build |
status_codes | OK, RETRY_REQUEST, STEP_FAILED, QUALITY_GATE_FAILURE, PIPELINE_COMPLETE |
retry_policy | Exponential backoff: base=30s, max=15min, 5 retries |
Overnight Execution
4-Stream Verification
Both phases employ four parallel verification streams (FORGE Code 06 pattern):Stream 1: Primary
Execute the methodology as specified. Produces the primary output.
Stream 2: Alternative
Independent approach using a different analytical framework.
Stream 3: Critical Review
Identify gaps, errors, unstated assumptions, blind spots.
Stream 4: Factual Verify
Re-verify claims against live data, flag discrepancies.
Traceability
Trace ID Requirement Rationale Verification DR-T-001 Zero human gating Eliminate overnight blocking on manual approval execution_manifest.yaml consumed by n8n without intervention DR-T-002 4-stream verification Reduce single-pass blind spots per FORGE Code 06 Synthesis section in all outputs DR-T-003 Epic-to-taskset traceability Audit finding → action mapping Traceability matrix in execution plan DR-T-004 A2A handoff contracts Formal agent-to-agent interface specification handoff_contracts.yaml validatable by JSON Schema DR-T-005 Exponential backoff Respect API rate limits, handle transient failures Retry policy in handoff contracts
Related Documentation
FORGE Pipeline Stages
FORGE Codes 01-06: ConOps through Churnery
Veritas Integration
How Veritas prompt library integrates with FORGE