Extensibility & Migration

Overview

PG Atlas is designed for iterative evolution beyond v0, supporting new signals, metrics, and scale without major rewrites. Extensibility focuses on modularity (abstracted layers, stable API) to enable community contributions and future features. Migration paths preserve data and logic when upgrading backends (e.g., from prototyping to production graph DB).

Guiding principles:

Modular architecture: Ingestion, storage, computation, API, and dashboard as loosely coupled components.
Stable public interface: API endpoints versioned; backward compatibility for consumers (dashboard, Tansu context).
Data portability: Export/import formats (Gremlin bulk, JSON dumps) for backend swaps.
Community-driven growth: PR-friendly extensions (new metrics via plugins, ingestion sources via config).

Long-term vision: Evolve into a comprehensive ecosystem observatory (on-chain usage, security scoring, predictive risk models) while remaining lightweight and transparent.

Key Extensibility Areas

New Metrics & Signals

On-chain integration: Add user-facing classic/Soroban tx stats for usage weighting (v1+).
Advanced activity scoring: Granular score replacing 4-value enum (commit recency, survey depth, deployment signals).
Security/Quality: CVE feeds, audit status attestation, test coverage from CI.
Plugin model: Metric modules (Python classes) registered via config; computed in batch pipeline.

Ingestion Extensions

New sources: GH API for issue/PR stats (reviewer diversity), on-chain manifests.
Versioned blast radius: Per-release dependency edges for vulnerability tracking.
External ecosystems: Broader crawling (e.g. ZK libs reused in Stellar).

Storage & Backend Migration

From v0 backend → migrate to JanusGraph (single-node BerkeleyDB → distributed Cassandra/Scylla). See Graph Scaling for future options.

API & Dashboard Growth

New endpoints: Transitive blast radius per version, trend histories.
Webhooks: Notify on metric changes (e.g., pony factor drop).
Dashboard modules: Community-submitted views (e.g., risk heatmap).

Risks & Mitigations

Tech debt: Strict abstraction enforcement; tests for provider swaps.
Scope creep: Prioritize via working group roadmap; v1 gated on 2026 outcomes.
Contributor friction: Clear extension guides (e.g., “add a metric” template).

Open Questions

Governance for new metric inclusion (working group vote vs. PR merge)?
Timeline for first migration (post-v0 stability, or earlier for Gremlin benefits)?
On-chain attestation for metrics (Soroban verifiable computations)?