Why this workflow exists

The average senior PM writes 40+ weekly status reports a year. At 60–90 minutes per report — collecting updates, reconciling RAG statuses, formatting for the exec audience — that's 60–90 hours of PM time annually per person. In a PMO of 20 PMs, that's a full FTE burned on formatting.

You can compress that to under 15 minutes per report with a repeatable AI workflow. Here's how, end to end.

The 4-step workflow

Step 1 — Assemble source material

Do NOT ask AI to write a status report from thin air. The output is only as good as what you feed it. Before you open Claude, ChatGPT, or Copilot, gather:

• Last week's Jira / Azure DevOps / Asana export (open, in progress, closed items)
• Workstream lead written updates (Slack, Teams, email)
• Meeting transcripts or notes from the past week (steering committee, working sessions)
• Current RAID log (risks, actions, issues, decisions)
• Any burn-down / burn-up / financials for the reporting period
• Last week's published status report (for continuity)

Drop all of it into a single document. This is your "source of truth" pack.

Step 2 — Use a locked template

The #1 mistake PMs make with AI status reports is asking for a "status update" and letting the model choose the structure. It won't match your PMO's format. It won't match what your steering committee expects. Every week the format will drift.

Fix this by locking a template. Here is a battle-tested one:

```

Weekly Status Report — [Project Name]

Reporting Period: [Start Date] – [End Date]
Reporter: [PM Name]

Executive Summary (4 sentences MAX)

[Where the project is, what changed this week, what needs steering attention]

Overall Status: [Green | Amber | Red]

Trend: [Improving | Steady | Deteriorating]

Milestones

| Milestone | Baseline Date | Forecast Date | Status | |---|---|---|---| | ... | ... | ... | ... |

Key Accomplishments This Week

- ...

Planned for Next Week

- ...

Risks & Issues Requiring Steering Attention

1. [Risk / Issue] — Owner: [Name] — Mitigation: [Action]

Decisions Required

1. [Decision needed] — Requested by: [Date]

Financials

Budget: $[X] | Actual to date: $[Y] | Forecast at completion: $[Z] | Variance: $[V] ```

Step 3 — Use a locked prompt

Same idea as the template. Don't improvise the prompt each week. Save the exact prompt you use and reuse it.

```
You are a senior program manager writing a weekly status report for
an executive steering committee.

RULES:
1. Use ONLY the source material provided below. Do NOT invent facts.
2. If a data point is missing, insert "PM to confirm" — never guess.
3. Match the attached template EXACTLY, including headings and order.
4. Keep the Executive Summary to 4 sentences maximum. Plain business
English, no marketing language.
5. In "Risks & Issues Requiring Steering Attention", include only
items that require action from the steering committee. Do NOT
include operational risks the team is already handling.
6. RAG rules:
- Green = on track, no material variance
- Amber = slippage < 10% OR risk requires mitigation
- Red = slippage >= 10% OR unresolved blocker

TEMPLATE:
[paste template from Step 2]

SOURCE MATERIAL:
[paste source-of-truth pack from Step 1]
```

Step 4 — Review, correct, send

The AI will get 80–90% right. Your job is to correct the last 10–20%. Specifically look for:

• RAG calls the model got wrong. Models tend to be too optimistic. If you know the project is Amber, don't let the model call it Green.
• Missing political context. The model doesn't know that the CFO is skeptical of this initiative. If steering needs to be pre-briefed, add that context.
• Overstatement of accomplishments. Trim confident language.
• Understatement of risks. Add real teeth to the risks section.

Then send. Total elapsed time: 10–20 minutes if you have the source pack ready.

What to do about tools that don't connect to your data

If you're on a locked-down enterprise stack where you can't paste Jira exports into a chat model, you have two options:

• Use your company's enterprise AI deployment. Claude for Enterprise, ChatGPT Enterprise, Microsoft 365 Copilot, or Gemini for Google Workspace all have zero-retention or in-tenant options that IT will usually approve for internal-classified data.
• Ask your CISO / privacy office what's approved for confidential material. Don't guess. Don't paste into a personal account.

Never paste confidential customer data, PHI, or regulated material into a consumer AI tool. Ever.

Common mistakes to avoid

• Asking for the report before assembling source material. The output is generic and useless.
• Not locking the template. Format drifts week to week and confuses your audience.
• Not locking the prompt. Output quality varies wildly.
• Skipping the review step. You are still the accountable PM. The steering committee is reading your name at the top.
• Using free consumer tools for confidential material. Don't.

What "great" looks like

The senior PMs who master this workflow report three things consistently:

1. Weekly status report cycle time drops from ~90 minutes to ~15 minutes.
2. Steering committee members comment on the improved clarity and consistency.
3. The PM has more time for the parts of the job that AI can't do — coaching, escalation, and delivery leadership.

That's the whole point. AI doesn't replace the PM. It replaces the parts of the PM's week that were never really adding value.

Next steps

If you want the full prompt library — status reports, meeting summaries, risk analysis, stakeholder comms — with a structured 12-module program on how to use AI as a senior delivery leader, see the Claude PM Pro course.

Or attend the free 60-minute masterclass to see the framework in action.