azdo-daily 1.0.0

Azure DevOps daily task automation with AI task breakdown

Azure DevOps Daily Task Automation

🤖 Automate daily task creation, activation, and completion logging in Azure DevOps with AI-powered task breakdown via Claude.

• AI Task Breakdown — Claude reads user stories and generates concrete development tasks
• Interactive Workflow — create → start → update → end pipeline
• Auto-linking — Tasks linked to parent stories, related across multiple stories
• Progress Tracking — Log hours, remaining estimates, completion notes
• Global CLI — Install once, use anywhere: azdo-daily configure

Installation

Install globally from this repository:

pip install .

Or install in development mode:

pip install -e .

Then use from anywhere:

azdo-daily configure
azdo-daily status

Configuration

On first run, configure your settings:

azdo-daily configure

Configuration stored in .config/settings.json (project-local, excluded from git).

Interactive prompts for:

Field	Description
org	Your Azure DevOps organisation name
project	Project name
pat	Personal Access Token — needs Work Items (Read & Write) scope
anthropic_api_key	From console.anthropic.com — for AI task breakdown
assigned_to	Your email/display name — used to filter stories
close_state	Done / Closed / Resolved (depends on your process template)

---

Project Structure

azdo_daily/
├── main.py       — Entry point: argument parsing + command dispatch
├── config.py     — Configuration: load/save/validate settings
├── state.py      — Daily state: load/save work items per day
├── ui.py         — Terminal UI: colors, prompts, formatters, selection helpers
├── azdo.py       — Azure DevOps API client: work item operations
├── ai.py         — Claude AI: task breakdown from stories
└── commands.py   — Command handlers: configure, create, start, update, end, status

Key Modules

azdo.py — Azure DevOps API client

• session() — authenticated HTTP session
• wit_base() — Work Item Tracking API base URL
• get_my_stories() — fetch assigned stories via WIQL
• create_task() — create new task linked to stories
• set_workitem_state() — change work item state
• resolve_task() — mark task complete with hours + comment
• partial_task() — log progress, keep task open
• _patch_workitem() — internal: apply JSON Patch operations
• add_comment() — internal: append work item comment

commands.py — CLI command handlers

• cmd_configure() — interactive setup
• cmd_create() — select stories → generate tasks → create in Azure DevOps
• cmd_start() — activate tasks (mark In Progress)
• cmd_update() — log progress on tasks
• cmd_end() — mark tasks done + auto-resolve stories
• cmd_status() — show today's stories and tasks

ui.py — Terminal UI helpers

• Color constants: R, B, DIM, GR, YL, BL, CY, RD, MG
• Output: ok(), err(), info(), warn(), hdr(), sep()
• Input: ask() (prompt with optional default)
• Formatters: print_stories(), print_tasks()
• Selection: parse_selection(), select_from_list() (handles "all")
• Conversion: float_or_none()

---

Daily Workflow

Setup (once)

azdo-daily configure

Morning — Create tasks

azdo-daily create

1. Fetches all active User Stories assigned to you
2. You select stories to work on (e.g. 1 or 1,3 or 2-4)
3. Choose task creation method:
   • AI auto-breakdown — Claude reads story and generates tasks
   • Manual entry — type tasks yourself
   • Both — AI suggests, you review/add/remove
4. If multiple stories: tasks created once, linked to all (child of #1, related to rest)
5. Tasks appear in Azure DevOps immediately

Activate tasks

azdo-daily start

Marks selected new tasks as "In Progress" and auto-activates linked stories.

Anytime — Check status

azdo-daily status

Shows active stories and all tasks with open/resolved counts.

Throughout day — Log progress (optional)

azdo-daily update

Log hours spent and remaining estimates on open tasks (keeps them "In Progress").

Evening — Mark complete

azdo-daily end

1. Lists all open tasks for today
2. You select which to complete (e.g. 1,3 or all)
3. For each: enter hours spent + optional closing note
4. Task marked done + hours logged
5. If all tasks for a story are done → story auto-resolved

---

State Files

Each day's data in state/YYYY-MM-DD.json:

{
  "stories": [
    {"id": 42, "title": "Implement auth"}
  ],
  "tasks": [
    {
      "id": 101,
      "title": "Add login endpoint",
      "url": "https://...",
      "active": true,
      "closed": false,
      "completed_hours": 2.5,
      "remaining_hours": 1.0,
      "story_ids": [42]
    }
  ]
}

---

Cron (optional)

Status reminder at 9 AM weekdays:

0 9 * * 1-5 azdo-daily status

start, update, and end commands are interactive — use cron for status only.

---

Troubleshooting

"Work item X does not exist"

Task IDs in state file no longer exist in Azure DevOps. Reset today's state:

rm ~/.config/azdo_daily/state/2026-04-29.json  # or today's date
azdo-daily create

"anthropic_api_key not set"

Run configure and enter Anthropic API key from console.anthropic.com:

azdo-daily configure

"Personal Access Token (PAT) invalid"

Regenerate PAT in Azure DevOps → User Settings → Personal Access Tokens. Must have Work Items (Read & Write) scope.

ModuleNotFoundError after install

Reinstall in development mode:

pip install -e .

---

Development

Install with dev tools:

pip install -e ".[dev]"

Format code:

black azdo_daily/

---

License

MIT — see LICENSE