Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →lawve-ai/risk-and-issues-manager-scott-margetts
RAID log methodology with decision extraction from emails and meeting notes.
Passed 17 of 22 test cases with the skill, and 14 of 22 without it.
and 9 more in the full report
6 fixed11 passed either way2 still failing3 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 6 test cases the skill fixed. Same model, same prompt, with and without the skill.
I am pasting below a recent email thread regarding our cross-border tech acquisition project (Project Horizon). Please review the full chain and extract all explicit and implicit choices, unexpected problems, potential uncertainties, underlying reliance points, and any indications of expanding or contracting scope. Format the output starting with a top summary of critical items requiring immediate partner action, followed by open item counts across each category, any scope deviation flags, and then detailed structured log entries for each identified item. [Email Chain Attached] From: Sarah Partner <[email protected]> To: M&A Team Date: October 12, 2023 Subject: Project Horizon - Scope and Structure updates Let's keep the French subsidiary out of the perimeter for now due to the labor council review timing. John, can you make sure we don't spend time on their tax history? Also, we're assuming the seller will provide audited Q3 statements by Friday. If local counsel in Munich doesn't confirm the merger clearance path by Tuesday, we'll need to rethink the closing timeline.
| Checked | Without skill | With skill |
|---|---|---|
| Uses ID prefixes I-, R-, A-, and D- for structured log entries. | Missed | Passed |
| Presents the IRAD log detail in the sequence Issues, Risks, Assumptions, and Decisions. | Missed | Passed |
| Includes an immediate action summary at the top referencing specific log IDs. | Missed | Passed |
The instructions your model reads when the skill loads.
Maintain a structured RAID log (Risks, Assumptions, Issues, Decisions) for legal matters and extract structured data from unstructured correspondence. This skill operates in two modes: RAID log maintenance (creating, updating, and managing the log) and decision extraction (identifying decisions, risks, assumptions, and scope-relevant information buried in email chains and meeting notes).
The decision extraction mode is the highest-value function. Decisions get made in email constantly — "let's leave Germany out for now," "we'll use the existing holdco," "client says they don't need a valuation" — and they're never formally recorded. Six months later when scope is disputed, budget is blown, or a workstream is stuck, nobody can find the email where the decision was made. This skill catches those decisions at the point they're made and logs them with the metadata needed to make them useful later.
Create, update, and manage the four components of the RAID log. This is the ongoing discipline — the log is a living document that's reviewed regularly and updated as the matter progresses.
The user pastes an email chain, meeting notes, or call notes. The skill scans for:
Output: structured entries ready to add to the RAID log, with flags for items that need follow-up or escalation.
The user describes a matter setup, and the skill identifies the key operational risks based on the matter type, jurisdictions, team structure, timeline, and commercial arrangements. This is proactive risk identification — catching risks at matter setup rather than discovering them mid-execution.
Ask the user what they need. Common triggers:
test cases passed, vs 14 / 22 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.