Dpdp compliance
Agent Skills for AI coding agents — starting with India's DPDP Act 2023 & Rules 2025 (audit, implement, cite). Cursor · Claude Code · Codex · more.
npx -y skills add KeshariPiyush24/agent-skills --skill dpdp-complianceAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 18 days oldThe repository was created 18 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Audits and implements India's Digital Personal Data Protection (DPDP) Act 2023 and DPDP Rules 2025 compliance. Use when auditing a codebase for DPDP gaps, implementing consent management, Consent Manager platforms, data principal rights APIs, breach notification, retention/erasure, verifiable parental consent, employment/HR (Section 7(i)) controls, or cookie/SDK gating, answering questions about the Act, Rules, penalties, Data Protection Board, or enforcement timelines, or generating privacy notices, consent records, breach drafts, DPIA checklists. Triggers include DPDP, Digital Personal Data Protection, India data privacy, data fiduciary, data principal, consent manager, Significant Data Fiduciary, MeitY. Do not use for GDPR-only, CCPA-only, or other non-Indian privacy regimes, and do not use as a substitute for professional legal advice.
The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.
SKILL.md
7.0 KB, as published. Nobody here has run it
DPDP Compliance Skill
Help software teams comply with India's Digital Personal Data Protection Act, 2023 (DPDP Act) and the Digital Personal Data Protection Rules, 2025 (DPDP Rules).
Legal status (state this honestly, never overstate)
- DPDP Act enacted 11 August 2023; DPDP Rules notified 13 November 2025 (G.S.R. 846(E), corrigenda G.S.R. 892(E) of 10 Dec 2025).
- Commencement is phased from 13 November 2025 (G.S.R. 843(E)):
- Now: Section 1(2), Section 2, Sections 18-26 (Board), Section 35, Sections 38-43 (incl. rule-making), Section 44(1)/(3); Rules 1, 2, 17-21.
- +12 months (13 Nov 2026): Section 6(9), Section 27(1)(d); Rule 4 and First Schedule (Consent Manager registration).
- +18 months (13 May 2027): Sections 3-5, 6(1)-(8) and (10), 7-17, Section 27 except 27(1)(d), Sections 28-34 (inquiry, appeals, ADR, voluntary undertaking, penalties), 36, 37, Section 44(2); Rules 3, 5-16, 22-23 and linked Schedules.
- Do not say appeals or penalties are in force before 13 May 2027.
- The Data Protection Board of India is established (HQ in the National Capital Region; G.S.R. 845(E) notifies four members — whether that count includes the Chairperson is not stated in the notification) — G.S.R. 844(E), 845(E).
- For exact dates and the penalty table, read
references/timelines-penalties.md.
Rules of engagement
- Always cite the specific Section (Act) or Rule (Rules 2025) for every claim or finding, e.g. "Section 8(7)" or "Rule 8, Third Schedule".
- Always include a disclaimer in reports and answers: "This is technical compliance guidance, not legal advice. Consult qualified counsel for legal decisions."
- Never invent obligations. If unsure, read the reference files below before answering.
- Respect the phased commencement: distinguish "in force now" vs "in force from 13 Nov 2026 / 13 May 2027" in every recommendation.
Workflows
Pick the workflow that matches the user's request. Read only the reference files that workflow needs (progressive disclosure).
1. Audit a codebase for DPDP compliance
When asked to audit, review, or assess a codebase/product for DPDP compliance:
- Read
references/audit-checklist.md(and skimexamples.mdExample 3 for report shape). - Complete Step 0 scope (role map, exemptions, SDF, Third Schedule, CM/HR/cookie flags).
- Map personal-data touchpoints — auth/signup, forms, DB schemas, analytics/SDK/cookies, logging, vendors, HR systems, exports.
- Work through every applicable checklist section (A–M); record status (compliant / gap / not-applicable / needs-human-review) with file:line evidence.
- Output a severity-ranked report (Critical / High / Medium / Low) citing Section/Rule, evidence, and a concrete fix. End with the disclaimer.
2. Implement compliance features
When asked to build or fix compliance functionality (consent flows, Consent Manager, rights APIs, retention jobs, breach handling, parental consent, HR/cookie controls):
- Read
references/implementation-guide.md. - Read the target repository's own conventions (framework, package manager, existing auth/user model) and follow them.
- Implement using the patterns in the guide; use
assets/consent-record-schema.mdfor consent data models. - In code comments and docs, cite the Section/Rule each feature satisfies.
3. Answer questions about the law
When asked what the Act/Rules say, definitions, obligations, penalties, timelines, or the Board:
- For Act questions read
references/dpdp-act-2023.md; for Rules questions readreferences/dpdp-rules-2025.md; for dates, penalties, and Board facts readreferences/timelines-penalties.md. Match tone ofexamples.mdExamples 1-2, 4-5. - Quote or closely paraphrase the actual text, cite Section/Rule numbers, and note whether the provision is in force yet.
- End with the not-legal-advice disclaimer.
4. Generate compliance artifacts
When asked to produce documents or templates:
- Privacy notice (Section 5, Rule 3): adapt
assets/privacy-notice-template.md. - Consent record data model (Section 6; notice versioning via Rule 3): adapt
assets/consent-record-schema.md. - Breach notification to Data Principals and the Board (Section 8(6), Rule 7): adapt
assets/breach-notification-template.md. - Data Protection Impact Assessment for Significant Data Fiduciaries (Section 10, Rule 13): adapt
assets/dpia-checklist.md.
Fill placeholders with the user's actual product/company details; never leave [PLACEHOLDER] text in delivered artifacts without flagging it.
Key concepts (quick reference)
- Data Principal: the individual the personal data is about (includes parents/guardians acting for children or persons with disability).
- Data Fiduciary: the entity that determines purpose and means of processing (analogous to a GDPR controller).
- Data Processor: processes on behalf of a Data Fiduciary under a valid contract.
- Significant Data Fiduciary (SDF): notified class with extra obligations — DPO in India, annual DPIA and audit, algorithmic due diligence, possible data-localisation restrictions.
- Consent Manager: registered platform letting Data Principals give, manage, review and withdraw consent (Rule 4, First Schedule).
- Applies to digital personal data processed in India, and outside India if connected to offering goods/services to Data Principals in India (Section 3). No extra "sensitive data" category — one tier of personal data.
Contents map
| Path | When to read |
|---|---|
references/audit-checklist.md | Audits (sections A–M) |
references/implementation-guide.md | Building features |
references/dpdp-act-2023.md | Act Q&A (paraphrase; verify India Code for legal decisions) |
references/dpdp-rules-2025.md | Rules/Schedules Q&A |
references/timelines-penalties.md | Dates, Board, penalties |
examples.md | Golden Q&A and audit finding shape |
assets/* | Notice, consent schema, breach, DPIA templates |