Skip to content

Q-agent Roadmap

Vision

Q-agent aims to become a reproducible, AI-compatible quantitative research workspace built around:

  • pipelines
  • notebooks
  • LEAN strategies
  • reusable signals
  • ObjectStore diagnostics
  • educational workflows

Recently Shipped

  • Docker / GHCR image (#20, #25). Public multi-architecture images bundle LEAN CLI, infrastructure pipelines, and marimo. Course users pin a release tag; :latest tracks main. See Docker.
  • Multi-arch Docker image — the published manifest supports both linux/amd64 and linux/arm64, so Apple Silicon hosts can pull natively. See Docker.
  • Workflow-first documentation site — mkdocs at https://wolfpackofone.github.io/Q-agent/ with linkcheck on every PR.
  • Personal-paths CI scanner — .github/workflows/secret-scan.yml blocks PRs that leak host home directories.
  • Pipeline maturity table — the data-pipeline catalog distinguishes stable, committed-data, and experimental pipelines.
  • Student contribution governance — fork-only workflow, CODEOWNERS, issue forms, expanded PR template, code of conduct, repository-policy checks, and private security reporting.
  • Required merge gates — stable Tests, Docs, Security, Repository policy, Docker, CodeQL, and Dependency review checks.

Near-Term Priorities

Documentation Hardening

  • close remaining docs-to-code consistency gaps
  • keep the Golden Path and ElectionIndustryBeta workflow synchronized
  • maintain pipeline maturity labels
  • improve agent workflow examples
  • keep onboarding commands copy-pasteable

Repo Hygiene / Course Operations

  • review the label and milestone taxonomy each term
  • add trusted TA reviewers before the instructor needs to submit a PR
  • keep course instructions pinned to a tested semantic release
  • periodically audit permissions, required checks, dependencies, and stale docs

Docker Followups

  • support lean backtest (local) inside the container — #27

Research Workflows

Current focus areas:

  • prediction markets
  • ETF sensitivity analysis
  • macro event research
  • volatility targeting
  • systematic portfolio construction
  • ObjectStore diagnostics

Tracked research feature issues:

  • #52 — passive instability research-to-strategy workflow
  • #62 — walk-forward analysis and bootstrap validation

Pipeline Expansion

Potential future pipelines:

  • FRED
  • SEC insider filings
  • options datasets
  • futures term structure data
  • macroeconomic indicators

Experimental pipelines already present but still maturing:

  • treasury_gov_rates
  • fixed_income
  • macro_rates

Agent-Native Development

Future goals:

  • reproducible Claude Code workflows
  • notebook generation workflows
  • research assistant tooling
  • AI-safe refactoring standards
  • project-level AGENTS.md conventions
  • graph-backed context packs and retrieval workflows

Graph-system roadmap issues:

  • #54 — hybrid property-graph-first architecture
  • #67 — stale-fact cleanup for merge-only re-ingest

Educational Goals

Q-agent is designed to support:

  • graduate quantitative finance education
  • systematic trading instruction
  • reproducible portfolio projects
  • AI-assisted software engineering workflows

Long-Term Direction

Potential future positioning:

  • canonical open-source educational quant workspace
  • AI-compatible quantitative development environment
  • reproducible LEAN research ecosystem
  • bridge between academic and operational quant workflows