Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →lawve-ai/serious-incident-reporting
Serious Incident Reporting (SIR) for high-risk AI systems under Article 73 EU AI Act (Regulation (EU) 2024/1689, KI-Verordnung). Assess, qualify, draft, and submit Article 73 Serious Incident Reporting notifications across the four Art. 3(49) harm categories: death or serious damage to health, critical infrastructure disruption, breach of fundamental rights under Union law, and serious property or environmental damage. Covers Art. 73 deadline buckets (2/10/15 days), market surveillance authority
Passed 21 of 22 test cases with the skill, and 18 of 22 without it.
and 1 more in the full report
4 fixed18 passed either way0 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
This benchmark measured an earlier version of the skill.
One of 3 test cases the skill fixed. Same model, same prompt, with and without the skill.
We host an automated candidate evaluation platform used across Germany. Yesterday we discovered a software flaw where applicants disclosing health-related workplace adjustments were systematically flagged and rejected without human review over the past three months. We need a formal assessment of our regulatory reporting duties under the EU AI Act. Please provide a scope and severity analysis memo, determine our mandatory submission timelines, outline which German supervisory bodies we must contact, and draft an initial regulatory incident notice along with a prioritized 24-hour internal action checklist.
| Checked | Without skill | With skill |
|---|---|---|
| Specifies a maximum reporting timeline of 2 days under Article 73(3). | Missed | Passed |
| Identifies Bundesnetzagentur as a primary German market surveillance contact point. | Passed | Passed |
| Includes an internal action checklist for the first 24 hours. | Passed | Passed |
The instructions your model reads when the skill loads.
This skill helps assess whether an event involving an AI system triggers serious-incident reporting duties under Article 73 EU AI Act, and then turns that legal analysis into a practical response plan, draft notification, and internal action list.
It is designed for providers, deployers, legal/compliance teams, product leads, and incident-response teams that need a usable process rather than abstract commentary.
Use this skill when the user wants to:
This workflow only applies if the system is actually a high-risk AI system under the EU AI Act.
That means:
If the system is only assumed to be high-risk because it sits near a regulated workflow, stop and confirm classification first. If Article 6(3) plausibly applies, incident reporting under Article 73 may not be triggered because the system is not in-scope as high-risk in the first place.
Depending on the request, produce some or all of the following:
test cases passed, vs 18 / 22 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.