Weekly status report template (free), plus the Claude skill that fills it in for me

By Viktor Szabó · Published 19 October 2026 · Updated 19 October 2026 · Checked against tool docs on 19 October 2026

If Friday afternoon is when you try to remember what happened this week, this post is for you. Below is a simple template you can copy, the fuller format my free Claude skill uses, a sample, and how to set it up.

Weekly status report template (copy it)

This is the short version. Copy it into a doc, an email or a Slack message. It works without any AI at all.

Weekly status report: week of DD Month YYYY

Shipped
- What got finished this week (one line each)

Blocked
- What's stuck: who owns it, what happens next

Key numbers
- The 3–5 numbers that matter, with last week's figure next to each

Next week's top 3
1.
2.
3.

The skill below uses a fuller version of the same idea.

Weekly status report format

This is the full structure the Weekly Ops Report Skill writes, in this order:

1. Title line: “Weekly Ops Report:” followed by the date range.
2. TL;DR: one or two sentences on how the week went.
3. Shipped: things that are done.
4. In progress: started but not finished.
5. Blocked: each item as item: owner, next action. If nobody owns it, it says “owner unclear”.
6. Key numbers: a small table with Metric, This week and Change.
7. Decisions made: what was actually decided.
8. Next week’s top 3: at most three priorities, based on blockers and unfinished work.
9. Missing data: anything mentioned without a value, or that couldn’t be checked.

The rules that make it work, with or without Claude: shipped means done, every blocker gets an owner (or an honest “owner unclear”), numbers come with their change, nothing gets invented, and the whole report stays under 300 words.

Weekly status report sample

Example (fictional business, not my real numbers). This sample was made by following the skill’s instructions step by step on the example input that ships in the download (shown further down). It uses the skill’s exact output format. A real run in Claude may word things a little differently.

The Friday team lunch was left out on purpose: the skill drops anything that isn’t work, a number or a decision.

Why a skill beats a saved prompt

A saved prompt lives in a notes file you have to find and paste, and copies drift. A skill is installed once in your Claude account. Anthropic describes skills as “written instructions, optionally with scripts and reference files, that Claude loads when the task you give it calls for them”, and names “your weekly report” as an example use (Claude docs). You ask, and Claude follows the same format every week.

Install the free Weekly Ops Report Skill

The skill is free on my store: Weekly Ops Report Skill. You get a ready-to-upload ZIP containing the skill folder, an example input and a short setup guide. No coding needed.

Upload the ZIP

According to Anthropic’s help center (Use skills in Claude):

1. Turn on code execution. On Free, Pro and Max plans, go to Settings → Capabilities and make sure Code execution and file creation is on. Skills need it.
2. Go to Customize → Skills.
3. Click +, then Create skill, then Upload a skill.
4. Upload weekly-ops-report.zip as it is. Don’t unzip it first.

The ZIP already has the weekly-ops-report folder at its root, which is what Claude expects.

Fallback: if you’ve already unzipped it, zip the weekly-ops-report folder itself, not the files inside it. Anthropic says the folder name must match the skill’s name (here, weekly-ops-report) and should sit at the root of the ZIP (How to create custom skills).

On Team and Enterprise plans, an owner controls skills, so ask them if Skills is missing.

Turn it on and test it

Turn the skill on in your skills list, open a new chat, paste the example input from the download and say:

“Make my weekly ops report.”

If Claude doesn’t pick it up, check the toggle and be more explicit: “Use my weekly ops report skill on these notes.”

An example: messy notes in, report out

This is the example input that ships in the ZIP (example-input.md). The skill also takes Slack dumps, task lists and CSV exports.

Example (fictional business, not my real numbers).

mon - call w/ acme, they want pilot in nov. payment bug fixed finally
tue hired Dana (support), starts 20th. landing page v2 live
wed - blocked on security questionnaire, waiting on legal?? tickets closed 46 (was 38)
thu 2 small accts left, reason: price. need to decide on annual plan
fri lunch w/ team. onboarding emails half done

That input became the sample report above. Note the messy parts:

• “waiting on legal??” isn’t treated as fact. The blocker says “owner unclear” and the question goes under Missing data.
• “tickets closed 46 (was 38)” becomes a number with its change.
• “need to decide on annual plan” isn’t listed as a decision, because nothing was decided. It becomes a priority for next week.

Make it automatic (optional)

Anthropic’s scheduled tasks run a task on a cadence using your connectors and skills, and weekly reports are one of the uses Anthropic lists. They’re on paid plans only (Pro, Max, Team and Enterprise). A simple setup is a Friday afternoon task that reads your notes doc and makes your weekly ops report for you to review before you send it.

If your inbox is where the week’s work actually happens, my free Gmail Inbox Triage Skill pairs well with this (here’s how I use it for inbox zero). It sorts recent email into Reply today, This week, Waiting on, FYI and Can archive, and changes nothing in Gmail unless you ask.

Keeping the numbers honest

The skill is written never to invent a number and to list anything it couldn’t verify under Missing data. On your side: paste the CSV or dashboard numbers rather than typing them from memory, check the Key numbers table against the source, and treat the Missing data line as your to-chase list. AI drafts, I check and send. That’s how I work on everything.

Want more than reports? The Claude + Connectors Starter Kit ($49) has 6 modules and 6 installable skills for marketing and admin work with Claude’s connectors. Three modules use real runs; the rest are clearly labelled illustrative walkthroughs.