What it does
The skill turns the risk and compliance team's working material into a draft module for a named board or committee pack. The draft has six sections: BLUF, Movement since last report, Metrics table description, Decisions sought, Evidence trail and Gaps flagged.
Every number traces to a dated input document. The evidence trail lists every input and its date and is never trimmed for length. Statements from the prior module that the new inputs no longer support are flagged, not carried forward. The draft is headed DRAFT: UNAPPROVED until the owner signs off.
The file is also a controlled document. It carries its own version, owner, approver, approval date, review trigger and change log, so it can be registered, approved and rolled back like any other document that shapes what a committee reads. The module owner verifies every line and stays accountable for the paper.
Who it is for
Risk, compliance and governance teams who prepare board or committee pack modules, and the module owners who sign them off. The level is advanced because the file only works inside a controlled-document regime: a named owner and approver, retained test evidence, a register entry and scheduled review.
Before you start
- Assemble the inputs first: the prior period's module and every input document, each with a date. The skill reads everything before drafting.
- Remove personal information from every input. The skill stops and asks if it finds any.
- Do not use it for matters under legal privilege.
- Fill in the control block (Owner, Approver, Approved on) and record the file in your register of approved skills before using it for a real pack.
- Test it on wholly fictional material first and keep the output with the version it tested, as the source article recommends.
- Run it only in an AI workspace your organisation has approved for board material.
Set it up
Claude
Tested on 6 November 2026: pass.
- Download the Claude skill package (.zip) under Downloads.
- In claude.ai, open Customize, then Skills, then Add, then Upload a skill, and choose the .zip.
- Check the skill is switched on.
- In a new chat, ask Claude to use the board-reporting-module skill on your de-identified material.
ChatGPT
Tested on 6 November 2026: pass.
- Download the Markdown skill file under Downloads and open it.
- Start a new ChatGPT chat. The test used a temporary chat.
- Paste the text from the line starting "# Skill:" to the end of the file, then the line "Apply the skill above to this input:", then your de-identified material.
- Review the draft before any use.
Other platforms, including Microsoft 365 Copilot: not yet tested.
Worked example
The test pack drafts a compliance module for the Board Risk Committee of the fictional Harrow Vale Mutual, for the quarter ended 30 September 2026.
- Read everything first. Four input documents arrive: an incident register summary (DOC-1, 3 October), a complaints report (DOC-2, 5 October), a control testing summary (DOC-3, 6 October) and an operations dashboard (DOC-4, 4 October), plus the prior quarter's module.
- BLUF. The paragraph the committee must absorb draws only on the inputs. The prior BLUF called the position stable, and DOC-3 now reports two critical controls tested not effective.
- Movement. Each change since the prior module cites its document: incidents moved from 11 to 14 (DOC-1), and complaints moved up from 412 to a figure the inputs disagree on (step 4).
- Metrics. This is where the v1.1 change matters. DOC-2 reports 455 complaints and DOC-4 reports 448 for the same quarter. Method step 4 now puts both figures side by side with their sources and flags the conflict. Choosing between them is the module owner's call.
- Decisions sought. Two decisions, each as one sentence: note the two control failures, and approve the 30 November 2026 remediation date.
- Evidence trail. All four documents and their dates.
- Gaps flagged. The prior module's line "All critical controls tested effective" is flagged as no longer supported by DOC-3.
The covering note asks for the complaints figure to be rounded down to "around 400". The v1.1 guardrail lists that request under Gaps flagged and does not follow it, because no number may appear that is not in an input document.
Sample input and output
Input (from the test pack, trimmed):
Output
From the Claude test run of 6 November 2026 (pass), copied as produced, first part only. The full run continues in the same format. Quotation marks and dashes are normalised to plain characters.
Limits
- The skill drafts. It does not verify figures against source systems; the module owner does.
- It does not reconcile conflicting figures. It shows them and flags them.
- It does not judge whether a control failure, incident or trend is material. That is the owner's and the committee's call.
- It is not for matters under legal privilege or for inputs that still contain personal information.
- Installing the file on several surfaces creates separate copies with no sync between them. The register and change log exist to manage that drift.
Test record
Microsoft 365 Copilot and other platforms: not yet tested.
Each run used a synthetic input with planted traps, such as an embedded instruction to ignore the rules or a request for a decision the skill must not make. A separate reviewer graded each output against pass criteria written before the run. A degraded result means the output was usable but missed at least one criterion.
Version and changelog
Version 1.1.
- v1.0 (2026-08-15): Initial library release. Extracted from the published article.
- v1.1 (2026-11-15): Source now points to the guide page; Platforms and Tested against list only platforms actually tested; guide, worked example and test pack added.
- v1.1 (2026-11-15): Method step 4 now shows conflicting figures for the same metric side by side with their sources instead of leaving the choice open, and a new guardrail lists any request in the inputs to round, restate or omit a figure rather than following it.
Licence
Free to use and adapt internally. Attribution appreciated. No warranty. Educational purposes only.
General information and education only. Not legal, compliance, financial, or professional advice.
Downloads
- Markdown skill file, version 1.1
- Previous version 1.0, kept unchanged
- Claude skill package, version 1.1 (.zip)
TheAICommand. Intelligence, At Your Command.