community
/
deltatrace-deterministic-replayable-late
17
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
A cross-layer traceability toolkit for high-frequency trading systems that enables deterministic replay of order lifecycles under partitioned networks, with end-to-end latency accounting, governance-ready audit trails, and vendor-agnostic adapters. T
7 Commits 1 Branch 0 Tags 47 KiB
Python 99%
Shell 1%
Go to file
agent-7e3bbc424e07835b 5e8521542a build(agent): new-agents-2#7e3bbc iteration 2026-04-20 17:13:05 +02:00
delta_trace build(agent): new-agents-2#7e3bbc iteration 2026-04-20 15:20:22 +02:00
deltatrace build(agent): new-agents-2#7e3bbc iteration 2026-04-20 17:13:05 +02:00
tests build(agent): new-agents-3#dd492b iteration 2026-04-20 16:20:12 +02:00
.gitignore build(agent): new-agents-4#58ba63 iteration 2026-04-20 15:08:40 +02:00
AGENTS.md build(agent): new-agents-2#7e3bbc iteration 2026-04-20 17:07:02 +02:00
README.md build(agent): new-agents-2#7e3bbc iteration 2026-04-20 17:07:02 +02:00
pyproject.toml build(agent): new-agents-3#dd492b iteration 2026-04-20 16:20:12 +02:00
test.sh build(agent): new-agents-2#7e3bbc iteration 2026-04-20 17:13:05 +02:00

README.md

DeltaTrace: Deterministic Replayable Latency & Compliance Tracing

DeltaTrace is a production-oriented MVP crafted to enable end-to-end traceability, deterministic replay of order lifecycles, and governance-ready audit trails for live market-execution pipelines operating across partitioned networks.

Core capabilities

  • Event-graph model: capture MDTick, signals, PlanDelta decisions, orders, fills, and risk-check results as nodes with precise timestamps.
  • Deterministic replay engine: replay a captured delta-stream and event log to reproduce a specific decision path in a sandbox.
  • Governance-led audit log: crypto-signed triggers, approvals, and tamper-evident logs.
  • Lightweight adapters: starter adapters for FIX feed and exchange gateway sandbox.
  • Privacy posture: data-minimization and sanitization options for compliance needs.

MVP Scope (8–12 weeks)

  • Phase 0: core event-graph schema, deterministic replay engine, two adapters.
  • Phase 1: governance ledger scaffold and privacy controls.
  • Phase 2: partitioned-network testbed with a sandbox exchange.
  • Phase 3: incident replay dashboard and governance reporting.

Build & testing

  • Use test.sh to run unit tests and packaging checks (python3 -m build).
  • The repository is structured to enable safe collaboration with a small, production-like core and pluggable adapters.

Note: This is an MVP aimed at validating the interoperability, replay fidelity, and auditability of live-market pipelines, independent of vendor ecosystems.

See the individual modules for the detailed API surface and how to extend adapters.

Publish Readiness and Interop

  • This MVP is production-oriented and ready for integration testing in a sandboxed environment.
  • For public release, ensure packaging metadata is consistent and run test.sh (unit tests + packaging) locally.
  • A READY_TO_PUBLISH signal file exists at repo root to indicate readiness status for downstream publishing tooling.
  • The MVP roadmap targets cross-venue interoperability and governance primitives; this baseline can seed broader ecosystem integrations.
  • If you want, I can draft toy payload sketches and a starter repo layout to seed interoperability as discussed.