Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →lawve-ai/eu-data-act-oliver-schmidt-prietz
Practitioner skill for advising on EU Regulation 2023/2854 (Data Act). Covers Chapters II-VII (IoT data access, mandatory B2B sharing, unfair contract terms, public-sector exceptional need, cloud switching, third-country governmental access) and Chapter VIII (interoperability and smart contracts, gate-only).
Passed 20 of 22 test cases with the skill, and 11 of 22 without it.
and 10 more in the full report
9 fixed11 passed either way2 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 9 test cases the skill fixed. Same model, same prompt, with and without the skill.
We are advising an enterprise customer negotiating a cloud migration contract with a European cloud infrastructure vendor. The vendor's draft contract includes standard terms regarding switching procedures, egress fees, and technical transition support. Prepare a legal review memo for senior counsel evaluating the draft switching provisions against EU Data Act requirements. Please structure the document by classifying vendor and customer legal roles, assessing the mandatory contractual terms for switching services, evaluating the transition fee rules, and noting any pending legislative updates that could affect these provisions.
| Checked | Without skill | With skill |
|---|---|---|
| Identifies the statutory switching fee abolition date under Article 29(2) as 12 January 2027. | Passed | Passed |
| Identifies Chapter VI as governing data processing service switching requirements. | Passed | Passed |
| Flags the Digital Omnibus proposal COM(2025) 833 final regarding Articles 25 and 31. | Missed | Passed |
The instructions your model reads when the skill loads.
This skill produces practitioner-grade analysis and drafting on Regulation (EU) 2023/2854 (the Data Act). It is calibrated for senior legal counsel, compliance officers, and product counsel working with the Data Act in client-facing or in-house contexts.
The skill's architectural anchor is role × chapter × stage. Every matter is positioned by identifying which Data Act roles the parties play (user, data holder, data recipient, third party, customer, provider, public sector body), which chapter of the regulation governs (II-VIII), and which stage of that chapter's process the matter is at (negotiation, request, response, refusal, enforcement). The anchor determines which references and templates load.
When invoked, read these files in order:
references/method/analysis-method.md — the seven-step cognitive flow the skill applies to every substantive matterreferences/method/house-style.md — output style and citation conventionsThen, based on the matter, load:
references/gotchas.md if the matter touches trade secrets, role mapping, "without undue delay" SLAs, the Art. 4(2) safety/security handbrake, gatekeeper exclusion, Ch VI custom-built carve-out, the sui generis right, or compensation direction. In practice this is most matters; the catalogue is short and worth reading on any substantive question.
The applicable gate file(s) in references/gates/:
gdpr-overlay.md whenever personal data is in scope, the user is a natural person, or the scenario involves terminal-equipment accesstrade-secrets-directive.md whenever any data is claimed or might be claimed as a trade secretdma-gatekeeper.md whenever a third party in an Art. 5 request could be a DMA-designated gatekeeper, or when downstream sharing under Art. 6(2)(c) is in scopesectoral-lex-specialis.md whenever the matter involves a regulated sector (automotive, medical devices, financial services, energy, AI, cybersecurity, agriculture, telecoms)member-state.md whenever the matter depends on Member State implementation (competent authority designation, complaint forum, penalties, dispute settlement)The applicable scenario card in references/scenarios/ (added in Phase 5).
Quote from the source files when stating what the regulation or FAQ says:
sources/regulation-2023-2854.md is the verbatim Data Actsources/faq-v1-4.md is the Commission FAQ (non-authoritative; frame as Commission interpretation)sources/digital-omnibus-amendments-tracker.md for current-law-vs-proposal discipline on affected provisionssources/mcts-sccs-recommendation-pointer.md and sources/vehicle-data-guidance-pointer.md for Commission soft-law instrumentsNever paraphrase the regulation from training data. Always quote from the source files. If the source file does not contain the needed passage, the analysis must not rely on it.
Before producing any output, the skill positions the matter on the anchor.
Role. For every entity in the scenario, identify Data Act role(s) and any concurrent GDPR role(s). The same entity can play multiple roles, and roles can shift across phases of the scenario. Role mapping is the most consequential analytical step; output that hides the mapping is unreliable. See references/method/analysis-method.md Step 3.
test cases passed, vs 11 / 22 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.