legal-setup
Skill1 starsChanged 4 days agoMITView on GitHub
Set up the legal compliance of a project or a website (global or web scope): writes the documents from the real data flows (privacy policy, mentions légales, terms CGU / CGV, cookie policy, accessibility statement, records of processing, retention policy, rights and breach procedures, credits, security.txt) after asking every fact it cannot know (identity, host, DPO, countries, languages - never invented), and implements the code (a CNIL-compliant consent banner blocking trackers until consent with refuse as easy as accept, data export and real erasure, purge jobs, form notices, unsubscribe, AI disclosure, accessibility fixes, footer links). Use whenever the user wants a project / site made legal or compliant, a privacy policy, mentions légales, CGU / CGV, a cookie banner, RGPD / GDPR setup, data deletion, an accessibility statement, or the fixes of a legal-check.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- TsukiNi22/skillsword for word, first seen Nov 16, 2026
- teodstr/garagetheir own version, first seen Nov 15, 2026
Other versions2 projects use a version of this.
Word for word · 1
- TsukiNi22/skills★ 1100% same
Edited · 0
None found.
Wrote their own version · 1
- teodstr/garage★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository70 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- ask-at-endSkill
- clarifySkill
- continue-workSkill
- exportSkill
- full-permissionSkill
- ideationSkill
- interactive-questionsSkill
- onboardSkill
- pauseSkill
- presentationSkill
- rtkSkill
- setup-contextSkill
- write-like-meSkill
- audit-bugsSkill
- audit-depsSkill
- audit-qualitySkill
- audit-securitySkill
- benchmarkSkill
- cmake-styleSkill
- cpp-classSkill
- cpp-commentsSkill
- cpp-projectSkill
- cpp-styleSkill
- better-explain-docSkill
- better-html-docSkill
- better-html-styleSkill
- better-protocol-docSkill
- docs-updateSkill
- explain-docSkill
- html-docSkill
- html-styleSkill
- protocol-docSkill
- readme-styleSkill
- reportSkill
- wiki-styleSkill
- ci-cdSkill
- git-commitSkill
- git-conventionsSkill
- pre-commitSkill
- legal-checkSkill
- licenseSkill
- libutils-exceptionSkill
- libutils-installSkill
- libutils-setupSkill
- libutilsSkill
- python-classSkill
- python-commentsSkill
- python-projectSkill
- python-styleSkill
- auditSkill
- cppSkill
- devSkill
- docSkill
- gitSkill
- legalSkill
- pythonSkill
- setup-projectSkill
- styleSkill
- setup-studioSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 19
Ask what only the user knows (skill interactive-questions, rounds of 1-4 questions): scope (global / web / both).
- Distinctive: line 77
Run the project's tests / build.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Legal setup (make it compliant)
Turn the obligations into documents and code for this project. The checklist is the one of legal-check
(../legal-check/references/global.md, web.md, sources.md): read it first. Not legal advice: binding texts (terms,
DPAs, a sensitive processing) are drafts to have reviewed by a lawyer, said once.
Request given with the skill: the text after /legal-setup if any (what to set up, a legal-check report).
1. Start from the facts
- A recent
legal-checkreport: take its findings (the fixes to do). None: run its inventory steps (§2 oflegal-check) first - documents describe the real data flows, never a generic template. - Ask what only the user knows (skill
interactive-questions, rounds of 1-4 questions): scope (global / web / both); the publisher's identity (company name, legal form, address, RCS / SIREN, VAT, share capital, publication director, contact email / phone) or an individual; the host (name, address, phone); DPO / privacy contact; the countries and the languages of the documents (the language of the users: French for a French site); B2C / B2B; what is sold; the mediator; the retention wishes; the analytics / marketing tools kept or dropped. - Never invent an identity, a number, a host, a mediator or a retention period: a
{{PLACEHOLDER}}stays, listed at the end. Confirm before overwriting an existing legal text (show the diff).
2. Documents (templates in templates/, filled and adapted, written in the language chosen)
| Document | When | Template |
| Privacy policy (art. 13-14) | personal data | privacy-policy.md |
| Legal notice / mentions légales (LCEN art. 6-III) | a website | legal-notice.md |
| Cookie policy + the list of the trackers | a site with any tracker | cookie-policy.md |
| Terms of use (CGU) / of sale (CGV) | accounts, user content / selling | terms.md |
| Accessibility statement (RGAA / EAA) | public sector, > €250 M, EAA services | accessibility-statement.md |
| Records of processing (art. 30) | personal data (always recommended) | records-of-processing.md |
| Retention policy + purge schedule | personal data | retention-policy.md |
| Data subject requests procedure (art. 12-22) | personal data | rights-procedure.md |
| Breach procedure + register (art. 33-34) | personal data | breach-procedure.md |
| Processors / sub-processors list + DPA checklist (art. 28) | any external service | in records-of-processing.md |
| security.txt (RFC 9116) / SECURITY.md (CRA vulnerability handling) | a site / a product | security.txt |
| Credits / open-source licenses (THIRD_PARTY_NOTICES, NOTICE contents, fonts, icons, images, maps) | anything distributed | audit-deps output |
| The project's LICENSE, SPDX ids, copyright headers | always | license |
Placed where the project keeps them (site pages / routes, docs/legal/, legal/), linked from the footer / the app | ||
| menu / the README; each page dated with a version, the old versions kept (terms). |
3. Code (the stack of the project, its style skills)
- Consent banner: nothing non-essential runs before consent (scripts blocked:
type="text/plain"+ loaded on consent, or a consent-mode loader); Accept all / Refuse all / Customize on the first layer, the same weight; purposes granular; choice stored ≤ 6 months then asked again; proof logged (time, version, choices); a "Cookies" link everywhere to change it; embeds behind click-to-load; fonts self-hosted. Prefer a maintained open-source CMP (tarteaucitron, Klaro, Orejime) configured for this, or a small own one when the trackers are few. - Audience measurement without consent only in the CNIL exempt configuration (first-party, IP truncated, 13 months / 25 months, opt-out offered) - else behind the banner.
- Rights: data export (JSON / CSV, art. 20) and deletion (art. 17) from the account and by request: cascade to every table, file, cache, index, processor (their deletion APIs), mailing list; backups' expiry documented; a confirmation step; a log of the requests (without the deleted data).
- Retention: purge / anonymisation jobs per data category (cron, scheduled task), log rotation, IP truncation.
- Forms: a short notice under each (who, why, basis, link), only needed fields, no pre-ticked opt-in, double opt-in
for newsletters, unsubscribe link and
List-Unsubscribein marketing emails. - Security basics (art. 32): password hashing (argon2 / bcrypt), HTTPS + HSTS, secure cookies, CSP, secrets out of the repository, rate limiting, admin 2FA.
- AI (AI Act art. 50, from 2 Aug 2026): a visible "you are talking to an AI" for chatbots, AI-generated content labelled (+ machine-readable marking when generating media).
- Accessibility: the fixes found (labels, alts, contrast, focus, keyboard, headings, lang), an axe / pa11y check in
the CI when there is one (
ci-cd), the statement linked from the footer. - E-commerce: withdrawal information + form, "order with obligation to pay" button, the ODR link removed, the mediator shown, easy subscription termination.
- Credits: an "About / Open-source licenses" screen or page generated from the dependencies (license comments kept in the bundles), the NOTICE contents of Apache-2.0 parts, attributions (CC-BY, OpenStreetMap...), the AGPL / GPL source offer when required.
- Language and mandatory mentions (
global.md§15): the documents and consumer information in the language of the market (French in France),langon pages and passages, prices with taxes, ads marked, product safety information on product listings. - Footer / menu: Legal notice · Privacy · Cookies (manage) · Terms · Accessibility · Credits, on every page.
4. Check and hand over
- Run the project's tests / build; for a site, load it again with an empty profile: no tracker before consent, refuse works, the links work, the pages render on a phone.
- Recap: what was written / changed (paths), the placeholders still to fill, what needs a lawyer / DPO review (terms,
DPAs, DPIA), the dates to watch (
sources.md⏱), and the next re-check (legal-check) recommended in 6-12 months. - Commits: the user's git rules and
git-conventions(nothing committed unless asked / allowed).
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

