Agent skill

Implementing Secure Multi Party Computation

by mukul975 in mukul975/Privacy-Data-Protection-Skills

Implementation guide for secure multi-party computation enabling privacy-preserving analytics across organizations.

Apache-2.0Auto-check passedLegal & Compliance

Install Implementing Secure Multi Party Computation

skills CLI
$ npx skills add mukul975/Privacy-Data-Protection-Skills --skill implementing-secure-multi-party-computation -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install mukul975/Privacy-Data-Protection-Skills implementing-secure-multi-party-computation --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/mukul975/Privacy-Data-Protection-Skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/privacy/implementing-secure-multi-party-computation .claude/skills/implementing-secure-multi-party-computation && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
implementing-secure-multi-party-computation
GitHub stars
295
Token cost
~2.2k tokens
SKILL.md length
845 words
Files
5 (incl. scripts, references, assets)
Skills in repo
278
Repo updated
First seen
Licence
Apache-2.0

At a glance

Implementation guide for secure multi-party computation enabling privacy-preserving analytics across organizations.

  • Works in 7 steps: Define the computation — Express the… → Select security model — Semi-honest if… → Select framework — Choose MP-SPDZ for… → …
  • Tasks that involve Privacy and GDPR
  • SKILL.md covers Overview, Core SMPC Techniques, SMPC Frameworks and Architecture for…, plus 4 more sections
  • Runs Python scripts from its folder

What it does

Implementing Secure Multi Party Computation is an agent skill from mukul975/Privacy-Data-Protection-Skills. Implementation guide for secure multi-party computation enabling privacy-preserving analytics across organizations. Covers secret sharing, garbled circuits, reference frameworks MP-SPDZ and CrypTen, practical deployment patterns, and GDPR alignment for joint controller analytics without revealing individual party inputs.

Its SKILL.md is about 2.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 7 other files, including scripts, reference files and assets (for example `assets/template.md`, `references/standards.md` and `references/workflows.md`).

It sits in Legal & Compliance, covering Privacy and GDPR. The repository describes itself as: 282+ structured privacy & data protection skills for AI agents. GDPR, CCPA, EU AI Act, HIPAA, LGPD, PIPL, DPDP Act. The licence is Apache-2.0.

When your agent uses it

  • Tasks that involve Privacy and GDPR

Example prompts

  • “/implementing-secure-multi-party-computation”

Requirements

  • Python 3

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

  1. Define the computation — Express the joint function as arithmetic or boolean operations.
  2. Select security model — Semi-honest if all parties are trusted (e.g., consortium members); malicious if adversarial behavior is possible.
  3. Select framework — Choose MP-SPDZ for flexibility, CrypTen for ML workloads, Sharemind for managed enterprise.
  4. Implement protocol — Write the computation in the framework's DSL, test on synthetic data.
  5. Benchmark — Measure end-to-end latency, communication volume, and memory usage.
  6. Deploy — Establish secure communication channels (mTLS), key distribution, and monitoring.
  7. Govern — Execute joint controller agreement (Art. 26), update Article 30 records, document in DPIA.

What it can do on your machine

Read from SKILL.md and the folder at commit 9b2ef9e. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 1 file in scripts/ (Python), which the agent can run.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Implementing Secure Multi Party Computation loads about 2.2k tokens when it runs, and up to ~4.5k if it reads all its reference files. Until then it costs about 92 tokens; SKILL.md has 845 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~92
When it runs · the whole SKILL.md, loaded when a task matches
~2.2k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~4.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from mukul975/Privacy-Data-Protection-Skills at commit 9b2ef9e, republished under its Apache-2.0 licence (© mukul975). 845 words, ~2,203 tokens.

Download SKILL.mdSave it as .claude/skills/implementing-secure-multi-party-computation/SKILL.md (or your agent's skills folder). This skill also uses 4 other files; get the full folder from GitHub.
name
implementing-secure-multi-party-computation
description
Implementation guide for secure multi-party computation enabling privacy-preserving analytics across organizations. Covers secret sharing, garbled circuits, reference frameworks MP-SPDZ and CrypTen, practical deployment patterns, and GDPR alignment for joint controller analytics without revealing individual party inputs.
license
Apache-2.0
metadata.author
mukul975
metadata.version
1.0
metadata.domain
privacy
metadata.subdomain
privacy-by-design
metadata.tags
secure-mpc, secret-sharing, garbled-circuits, mp-spdz, crypten

Implementing Secure Multi-Party Computation

Overview

Secure Multi-Party Computation (SMPC) enables multiple parties to jointly compute a function over their combined inputs while keeping each party's individual input private. No party learns anything beyond the output of the computation and what can be inferred from their own input and the output.

SMPC supports GDPR Article 5(1)(c) data minimization by eliminating the need to centralize data, Article 25(1) data protection by design by building privacy into the computation architecture, and Article 26 joint controller arrangements by enabling collaborative analytics without data sharing.

Core SMPC Techniques

Shamir Secret Sharing

Shamir's Secret Sharing (1979) splits a secret value into n shares such that any t shares can reconstruct the secret (threshold t-out-of-n), but fewer than t shares reveal no information.

Properties:

  • Information-theoretic security (unbreakable regardless of computational power)
  • Supports addition of shared values without communication
  • Multiplication requires an interactive protocol (Beaver triples or resharing)

Application: Splitting personal data across multiple servers such that no single server (or coalition below threshold) can reconstruct the data.

Garbled Circuits

Yao's Garbled Circuits (1986) enable two-party computation. One party (the garbler) encrypts a boolean circuit; the other party (the evaluator) evaluates the encrypted circuit without learning intermediate values.

Properties:

  • Constant-round protocol (efficient for high-latency networks)
  • Communication cost proportional to circuit size
  • Best for two-party computation with complex boolean functions
Oblivious Transfer

Oblivious Transfer (OT) is a protocol where a sender has multiple messages and a receiver selects one message to receive, without the sender learning which message was selected and without the receiver learning the other messages.

Role in SMPC: OT is the foundational building block for garbled circuit evaluation and is used for input wire labels in Yao's protocol.

SMPC Frameworks

FrameworkProtocolLanguageSecurity ModelBest For
MP-SPDZMultiple (SPDZ, MASCOT, semi-honest, malicious)Python-like DSLSemi-honest and maliciousResearch and prototyping with multiple security models
CrypTenSecret sharing (2-party and 3-party)Python (PyTorch)Semi-honestML inference and training on shared data
MOTIONGMW, BMR, arithmetic/boolean sharingC++Semi-honestHigh-performance 2+ party computation
ABY/ABY3Arithmetic, Boolean, Yao sharingC++Semi-honest (2-party ABY, 3-party ABY3)Mixed-protocol computation (switching between share types)
SharemindAdditive secret sharing (3-party)SecreC (DSL)Semi-honest (honest majority)Enterprise deployment with managed infrastructure
MP-SPDZ Protocol Selection
ProtocolSecurity ModelPartiesPreprocessingOnline Performance
SPDZ/MASCOTMalicious (dishonest majority)2+Heavy (OT-based)Fast online phase
Semi2kSemi-honest (dishonest majority)2+ModerateFast
ShamirSemi-honest (honest majority)3+LightVery fast
Rep3Semi-honest (honest majority)3NoneVery fast
MASCOTMalicious (dishonest majority)2+OT-basedModerate
Yao's GCSemi-honest2GarblingFast evaluation

Architecture for GDPR-Compliant SMPC

┌─────────────────────────────────────────────────────────────┐
│                  SMPC Computation Coordinator                │
│              (orchestrates protocol execution)               │
└──────┬─────────────────┬─────────────────┬──────────────────┘
       │                 │                 │
┌──────▼──────┐  ┌───────▼──────┐  ┌──────▼──────┐
│  Party A    │  │  Party B     │  │  Party C    │
│  (Bank)     │  │  (Insurer)   │  │  (Retailer) │
│             │  │              │  │             │
│ Input: x_A  │  │ Input: x_B   │  │ Input: x_C  │
│ Share: [x_A]│  │ Share: [x_B] │  │ Share: [x_C]│
│             │  │              │  │             │
│ Compute on  │  │ Compute on   │  │ Compute on  │
│ local share │  │ local share  │  │ local share │
└──────┬──────┘  └───────┬──────┘  └──────┬──────┘
       │                 │                 │
       └────────────────►│◄────────────────┘
                         │
                  ┌──────▼──────┐
                  │  Output:    │
                  │  f(x_A,x_B, │
                  │    x_C)     │
                  │  (revealed  │
                  │   to all)   │
                  └─────────────┘

Each party learns ONLY the final output f(x_A, x_B, x_C),
NOT the individual inputs of other parties.

Practical Deployment Patterns

Pattern 1: Private Set Intersection (PSI)

Two parties determine which records they have in common without revealing records unique to either party.

Use case: A bank and an insurer identify shared customers for a joint risk assessment without revealing their full customer lists.

StepActionPrivacy Guarantee
1Each party hashes their customer identifiers with a shared keyIndividual identifiers not revealed
2Parties engage in PSI protocol (DH-based or OT-based)Only intersection elements revealed
3For matched records, proceed with joint computation on shared dataNon-matched records remain private
Show full SKILL.md (323 more words)Show less
Pattern 2: Private Aggregation

Multiple parties compute aggregate statistics (sum, mean, count) over their combined data without revealing individual contributions.

Use case: Three regional offices of Prism Data Systems AG compute total headcount and average salary without revealing per-office figures.

Pattern 3: Private ML Training

Multiple parties jointly train a machine learning model using SMPC to protect training data during the process.

Use case: Using CrypTen, two hospitals jointly train a logistic regression model on encrypted patient features.

Performance Considerations

FactorImpactMitigation
Network latencySMPC is communication-intensive; high latency increases total timeCo-locate computation nodes; use constant-round protocols (Yao) for WAN
Data sizeCommunication scales with input sizePre-aggregate locally where possible; use sketching
Circuit complexityDeeper circuits require more communication rounds (for GMW)Optimize circuit depth; use Yao for constant rounds
Number of partiesMore parties = more communication pairsUse 3-party protocols with honest majority for efficiency
Security modelMalicious security is 10-100x slower than semi-honestUse semi-honest for trusted consortium members; reserve malicious for adversarial settings

Implementation Workflow

  1. Define the computation — Express the joint function as arithmetic or boolean operations.
  2. Select security model — Semi-honest if all parties are trusted (e.g., consortium members); malicious if adversarial behavior is possible.
  3. Select framework — Choose MP-SPDZ for flexibility, CrypTen for ML workloads, Sharemind for managed enterprise.
  4. Implement protocol — Write the computation in the framework's DSL, test on synthetic data.
  5. Benchmark — Measure end-to-end latency, communication volume, and memory usage.
  6. Deploy — Establish secure communication channels (mTLS), key distribution, and monitoring.
  7. Govern — Execute joint controller agreement (Art. 26), update Article 30 records, document in DPIA.

Key Regulatory References

  • GDPR Article 5(1)(c) — Data minimization (no data centralization needed)
  • GDPR Article 25(1) — Data protection by design
  • GDPR Article 26 — Joint controller arrangements for SMPC consortia
  • GDPR Article 28 — Processor obligations (SMPC coordinator role)
  • GDPR Article 32(1) — Security measures
  • GDPR Recital 78 — Technical measures for data protection
  • ENISA Report: Data Protection Engineering (2022)

© mukul975, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 4 other files (scripts, references, assets) in skills/privacy/implementing-secure-multi-party-computation of mukul975/Privacy-Data-Protection-Skills.

  • SKILL.md
  • assets/template.md
  • references/standards.md
  • references/workflows.md
  • scripts/process.py

Open the folder on GitHubat commit 9b2ef9e

Compare with similar skills

Implementing Secure Multi Party Computation next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Implementing Secure Multi Party Computation compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Implementing Secure Multi Party Computation this skillmukul975/Privacy-Data-Protection-Skills295—~2.2kAutomated safety check: PassApache-2.0
C15tc15t/c15t1.9k1 repos~1.6kAutomated safety check: PassApache-2.0
HIPAA Safe Harbor Coverage Auditmaziyarpanahi/openmed5.5k—~1.7kAutomated safety check: PassApache-2.0
Korean Privacy Termskimlawtech/korean-privacy-terms586—~2.9kAutomated safety check: PassApache-2.0
Gdpr ComplianceSushegaad/Claude-Skills-Governance-Risk-and-Compliance9421 repos~3.9kAutomated safety check: PassMIT
Hipaa ComplianceSushegaad/Claude-Skills-Governance-Risk-and-Compliance9421 repos~2.3kAutomated safety check: PassMIT

Similar skills

  • C15t

    c15t/c15t

    Work with c15t consent management docs, APIs, and integrations for Next.js, React, and JavaScript.

    1.9k GitHub starsUsed in 1 repo~1.6k tokens
    Legal & ComplianceAuto-check passed
  • Checks OpenMed de-identified clinical text against the 18 HIPAA Safe Harbor identifier categories and reports gaps and residual re-identification risk.

    5.5k GitHub stars~1.7k tokensUpdated today
    Legal & ComplianceAuto-check passed
  • Korean Privacy Terms

    kimlawtech/korean-privacy-terms

    처리방침·이용약관 자동 생성 스킬 패키지 (v4.0). An agent skill from kimlawtech/korean-privacy-terms.

    586 GitHub stars~2.9k tokensUpdated 1 mo ago
    Legal & ComplianceAuto-check passed
  • Gdpr Compliance

    Sushegaad/Claude-Skills-Governance-Risk-and-Compliance

    Expert GDPR compliance assistant covering all four core workflows: (1) auditing code and systems for GDPR violations, (2) drafting GDPR-compliant documents such as privacy policies, Data Processing…

    942 GitHub starsUsed in 1 repo~3.9k tokens
    Legal & ComplianceAuto-check passed
  • Hipaa Compliance

    Sushegaad/Claude-Skills-Governance-Risk-and-Compliance

    Expert HIPAA compliance assistant for healthcare and software contexts.

    942 GitHub starsUsed in 1 repo~2.3k tokens
    Legal & ComplianceAuto-check passed
  • Pii Contract Analyze

    gregmos/PII-Shield

    Universal legal document processor with PII anonymization. An agent skill from gregmos/PII-Shield.

    149 GitHub stars~8.9k tokensUpdated 3 mo ago
    Legal & ComplianceAuto-check: notes

More from mukul975/Privacy-Data-Protection-Skills

All 278 skills in this repo
  • Age Gating Services

    mukul975/Privacy-Data-Protection-Skills

    Implements age-gating mechanisms for online services to restrict access based on user age.

    295 GitHub stars~3.7k tokensUpdated 6 mo ago
    Auto-check passed
  • AI Data Retention

    mukul975/Privacy-Data-Protection-Skills

    Manages AI model retention and machine unlearning requirements.

    295 GitHub stars~1.9k tokensUpdated 6 mo ago
    Auto-check passed
  • Dpia Mitigation Plan

    mukul975/Privacy-Data-Protection-Skills

    Structures risk mitigation planning and residual risk tracking for Data Protection Impact Assessments under GDPR Article 35(7)(d).

    295 GitHub stars~846 tokensUpdated 6 mo ago
    Auto-check passed
  • Gdpr Accountability

    mukul975/Privacy-Data-Protection-Skills

    Guides implementation of the GDPR accountability principle under Articles 5(2) and 24, including documentation requirements for policies, DPIAs, RoPA, training records, and breach logs.

    295 GitHub stars~1.9k tokensUpdated 6 mo ago
    Auto-check passed
  • Pia Threshold Screening

    mukul975/Privacy-Data-Protection-Skills

    Conducts pre-DPIA threshold screening to determine whether a full Data Protection Impact Assessment is required under GDPR Article 35.

    295 GitHub stars~880 tokensUpdated 6 mo ago
    Auto-check passed
  • Retention Schedule

    mukul975/Privacy-Data-Protection-Skills

    Designs and implements data retention schedules compliant with GDPR Article 5(1)(e) storage limitation principle.

    295 GitHub stars~3.3k tokensUpdated 6 mo ago
    Auto-check passed

Questions about Implementing Secure Multi Party Computation

What does Implementing Secure Multi Party Computation do?

Implementation guide for secure multi-party computation enabling privacy-preserving analytics across organizations. Implementing Secure Multi Party Computation is an agent skill from mukul975/Privacy-Data-Protection-Skills. Implementation guide for secure multi-party computation enabling privacy-preserving analytics across organizations.

When should I use Implementing Secure Multi Party Computation?

Implementing Secure Multi Party Computation fits situations like: tasks that involve Privacy and GDPR.

How do I install Implementing Secure Multi Party Computation in Claude Code?

Run `npx skills add mukul975/Privacy-Data-Protection-Skills --skill implementing-secure-multi-party-computation -a claude-code`. Or copy the skill folder (skills/privacy/implementing-secure-multi-party-computation in mukul975/Privacy-Data-Protection-Skills) into .claude/skills/implementing-secure-multi-party-computation in your project. Claude Code loads it when a task matches its description.

How do I install Implementing Secure Multi Party Computation in Codex?

Run `npx skills add mukul975/Privacy-Data-Protection-Skills --skill implementing-secure-multi-party-computation -a codex`. Or copy the skill folder (skills/privacy/implementing-secure-multi-party-computation in mukul975/Privacy-Data-Protection-Skills) into .agents/skills/implementing-secure-multi-party-computation in your project. Codex loads it when a task matches its description.

Can I use Implementing Secure Multi Party Computation in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add mukul975/Privacy-Data-Protection-Skills --skill implementing-secure-multi-party-computation -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/implementing-secure-multi-party-computation, .gemini/skills/implementing-secure-multi-party-computation, .github/skills/implementing-secure-multi-party-computation and .opencode/skills/implementing-secure-multi-party-computation in your project.

What does Implementing Secure Multi Party Computation need to run?

Going by SKILL.md and its folder, Implementing Secure Multi Party Computation needs Python for the scripts in its folder. Our summary lists: Python 3.

Does Implementing Secure Multi Party Computation access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Implementing Secure Multi Party Computation safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Implementing Secure Multi Party Computation use?

Implementing Secure Multi Party Computation is published under the Apache-2.0 licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Implementing Secure Multi Party Computation use?

About 2.2k tokens (SKILL.md is roughly 8.8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2.3k tokens, read only when the agent opens those files.

What are the alternatives to Implementing Secure Multi Party Computation?

Skills that share tags, products or a category with Implementing Secure Multi Party Computation: C15t (c15t/c15t, 1.9k stars), HIPAA Safe Harbor Coverage Audit (maziyarpanahi/openmed, 5.5k stars), Korean Privacy Terms (kimlawtech/korean-privacy-terms, 586 stars) and Gdpr Compliance (Sushegaad/Claude-Skills-Governance-Risk-and-Compliance, 942 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Implementing Secure Multi Party Computation?

mukul975 (a GitHub user) maintains it in mukul975/Privacy-Data-Protection-Skills, which has 295 GitHub stars. The repository holds 278 skills in this directory. The repository was last updated on March 16, 2026.

Source: mukul975/Privacy-Data-Protection-Skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.