Hyper-Agent

Shadow

Your AI test architect

AI brain Second brain Clarity layer
Repository Full README Sample brief Nightly Headquarters

See the output

Morning brief

What you get after a run: bullets, suggested focus, and provenance (UTC time + sources). Yours can combine Azure DevOps open bugs with files under data/.

shadow — morning brief

Morning brief

  • Defects: 24 total, 4 critical/high, 18 not closed.
  • Top severity: ado-4412 — Payment API timeout under load (staging)…
  • Latest test run: Nightly regression — Passed. Passed: 312, Failed: 2

Suggested focus today: Review the two failing nightly cases with QA; confirm fix target for ado-4412 before delivery sync.

Sources: Azure DevOps (open Bugs), test_runs.json · As of illustrative snapshot

Flow

How it works

Three steps from the systems you already use to a brief you can defend in the room.

1

Connect

ADO, files, and (soon) calendar—signal from systems you already trust.

2

Compress

Briefs and stakeholder-ready framing—with provenance—not another dashboard to babysit.

3

Decide

You lead QA, delivery, and risk; Shadow never replaces your signature.

Positioning

Intelligence layer, not a replacement

Shadow sits on your stack and your rituals—then hands you a clear picture and wording you still own.

Why Shadow

Shadow sits on top of Azure DevOps, Outlook, and how you already run quality—not in place of them. It ingests work items and exports, turns them into a morning brief and (roadmap) meeting prep, and frees attention for judgment and relationships.

Hyper-Agent is the open project that implements Shadow; this site is the front door.

Live today

  • CLI & API morning brief
  • Optional live Azure DevOps open bugs
  • File-based test run summaries
  • Timestamped saves under output/briefs/
  • LLM polish when you enable it

Scope

Six capability areas

The roadmap is organized around how Test Managers actually spend their week—not feature sprawl.

01
Daily orchestration
Brief · priorities · meeting prep
02
QA team oversight
Commitments · capacity · consistency
03
Delivery collaboration
Scope ↔ tests · readiness · comms
04
Decision support
Go/no-go evidence · impact
05
Governance
Standards · patterns
06
Second brain
Status on demand · your prefs

Resources

Documentation

Vision, diagrams, build plan, ADO + Outlook, deploy, and next steps—all in the repo.

Vision Diagrams Build plan ADO & Outlook Deploy API Next steps

Nightly run · Morning-ready

Nightly Headquarters

Your daily command center: one HTML page with the same snapshot as the morning brief—defects, test runs, sources, suggested focus, and optional quick links to ADO or dashboards. Run it on your machine before standup, hit the API from a deployed service, or let GitHub Actions build it every night so you open a fresh artifact in the morning.

Open sample Headquarters Full-page operations HQ preview (new tab): portfolio statuses, sprint execution, automation health — illustrative only.

This machine

From the repo root, generate output/headquarters/latest.html plus a timestamped copy and the markdown brief under output/briefs/.

pip install -r requirements.txt
python src/main.py headquarters

REST API

Same dashboard over HTTP—useful when Shadow runs on a small host or your network. Optional ?persist=1 writes files if the process can reach the repo path.

uvicorn src.api:app --reload
→ GET /headquarters.html

GitHub Actions

Workflow Nightly Headquarters runs on a schedule (and manually). Download the latest artifact and open latest.html inside—no local Python required.

Open workflow runs

Configure quick links and title under headquarters in config/config.yaml. Add repository secret AZDO_PAT for live Azure DevOps bugs in the snapshot.

Run your first brief

Clone the repo, install dependencies, then generate a brief (and a timestamped file under output/briefs/).

pip install -r requirements.txt
python src/main.py brief

REST API: uvicorn src.api:app --reload → /brief.md