Today's Agent Skill: Auto-Draft Your Weekly Status Report

What It Does

Writing a weekly status report takes 20–40 minutes of your Friday afternoon — time spent mostly translating your calendar and task list into prose your manager will skim in 30 seconds. The painful part isn't the work itself; it's excavating what you actually did from a week of scattered notes, Slack threads, and half-closed tabs. An agent can do that excavation and first draft in under two minutes.

How It Works

The skill pulls raw inputs — a bullet list of completed tasks, blockers, and upcoming priorities — and structures them into a consistent four-section format: wins this week, blockers encountered, metrics moved, and top three focuses for next week. It normalizes vague entries into clean past-tense statements wherever possible, and flags items that need a number or outcome filled in before the report is sent.

How to Deploy It

Add the SKILL.md content below to your agent's skills directory and trigger it by pasting your raw task dump with the phrase 'draft my status report'. Works with any Claude Code-compatible agent that reads SKILL.md files from its skills folder at session start.

SKILL.md — Ready to Deploy

# weekly-status-draft

## Description
Transforms a raw bullet dump of weekly activity into a polished, manager-ready status report. Structures output into four sections: Wins, Blockers, Metrics Moved, and Next Week Priorities. Flags incomplete entries (missing outcomes or numbers) rather than guessing.

## Trigger
User message contains 'draft my status report', 'write my status update', or 'status report from these notes'.

## Input
- A raw list of tasks, notes, or bullet points from the user's week
- Optional: recipient name or team context (e.g., 'for my engineering manager')
- Optional: report period (defaults to current week if omitted)

## Steps
1. Parse the raw input and classify each item as: completed task, in-progress item, blocker, metric or outcome, or upcoming work.
2. For completed tasks, rewrite each as a crisp past-tense statement. If a quantifiable outcome is implied but not stated (e.g., 'fixed the login bug'), add a [NEEDS OUTCOME] flag inline rather than inventing a number.
3. Group all items into four labeled sections: **Wins This Week**, **Blockers / Risks**, **Metrics Moved**, **Top 3 Next Week**.
4. Under Metrics Moved, surface any numbers mentioned in the in

Copy the full SKILL.md and drop it into your agent's skills directory to activate this skill.

← Back to All Posts | Home — NJ AI Tools