AI Skills Marketplace

Design RESTful APIs from plain-language requirements. Produces OpenAPI 3.0-compatible endpoint definitions with request/response schemas, status codes, auth patterns, and pagination strategy.

Write technical blog posts from an outline or topic. Produces SEO-optimized, developer-friendly content with code examples, analogies, and a clear narrative arc.

Multi-pass code review analyzing logic errors, security issues, performance bottlenecks, and style consistency. Returns confidence-scored findings with actionable fix suggestions.

Write conventional commit messages from git diffs. Follows Conventional Commits spec with type, scope, subject, body, and breaking change footers. Understands code context to write meaningful messages.

Save project decisions, architecture notes, and lessons learned to structured context files. Enables persistent memory across AI coding sessions in Mistral Vibe and other tools.

Audit package manifests for outdated dependencies, known vulnerability patterns, license risks, and bloat. Works with package.json, requirements.txt, Gemfile, go.mod, and Cargo.toml.

Generate API documentation, README files, architecture decision records (ADRs), and changelogs from code analysis. Supports multiple documentation formats and follows established conventions.

Draft professional emails from bullet points or intent. Handles sales outreach, follow-ups, internal announcements, escalations, and partnership proposals with appropriate tone calibration.

Analyze code, data flows, and processes for GDPR compliance gaps. Identifies personal data handling, missing consent mechanisms, retention policy violations, and data subject rights gaps.

Write structured incident reports and postmortems from raw notes or a timeline. Produces blameless postmortems with impact analysis, root cause, timeline, and action items following SRE best practices.

Audit open source dependency licenses for compatibility with your project's license and commercial use restrictions. Flags copyleft, viral, and unknown licenses with risk assessment.

Build, explain, and test regular expressions from natural language descriptions. Provides the pattern, step-by-step breakdown, edge cases, and test strings.

Generate human-readable release notes from git commit history or a list of changes. Groups changes by type, highlights breaking changes, and calibrates technical depth to audience.

Extract structured user stories, acceptance criteria, and technical requirements from business documents, meeting notes, or feature descriptions. Outputs INVEST-compliant user stories with Gherkin acceptance criteria.

Generate operational runbooks for services, deployments, and incident response. Produces step-by-step procedures with verification steps, rollback instructions, and escalation paths.

Scan code for OWASP Top 10 vulnerabilities, dependency risks, and security anti-patterns. Outputs a structured report with severity levels and fix suggestions.

Break down features into structured implementation tasks with file paths, test requirements, complexity estimates, and dependency ordering. The planning tool Mistral Vibe is missing.

Build SQL queries from natural language questions. Supports SELECT, aggregations, JOINs, window functions, and CTEs. Explains each query component and suggests indexes.

Generate unit and integration tests for any function or module. Produces edge case coverage, mocks for dependencies, and test descriptions following AAA (Arrange-Act-Assert) pattern.

Generate structured changelogs from PRDs, feature lists, or informal change descriptions. Follows Keep a Changelog format and categorizes changes by type.

Translate content with cultural localization awareness. Adapts idioms, tone, date formats, and cultural references for the target locale — not just word-for-word translation.

Analyze CSV, JSON, or tabular data to surface key insights, statistics, trends, and anomalies. Returns structured findings with actionable recommendations.

Generate Mermaid or PlantUML diagrams from plain-language descriptions. Supports flowcharts, sequence diagrams, class diagrams, ER diagrams, and Gantt charts.

Generate targeted technical interview questions and detailed model answers for any engineering role and seniority level.

Extract decisions, action items, blockers, and key discussion points from meeting transcripts or notes. Returns a structured summary ready to share with attendees.

Analyze and rewrite prompts for optimal performance with Mistral AI models. Applies role-setting, output format definition, and instruction clarity best practices.

Reviews code for issues then generates a comprehensive test suite targeting the identified edge cases. Two-step pipeline: code-reviewer → test-generator.

End-to-end fine-tuning pipeline: formats raw training examples into Mistral JSONL, then generates the complete fine-tuning job submission code.

Before/after evaluation workflow: runs code-reviewer on a sample, then generates a comparison report to measure output quality improvement after fine-tuning.

Semantic search and Q&A over documentation, PDFs, or markdown files. Paste document content and ask questions — get cited answers with the exact passage that supports them.

Ask questions about any codebase. Paste code files, architecture docs, or README content and get precise answers with source references. Uses deterministic local semantic indexing across your code.

Check code, processes, or system designs against company policies, regulatory requirements, or internal standards. Paste your policies and the artifact to check — get a compliance gap report with citations.

End-to-end data analysis agent. Analyzes raw data, constructs the optimal SQL query to extract insights, then synthesizes a business narrative from both outputs.

Guides incident commanders through a structured response workflow. Uses incident-reporter to draft the postmortem and runbook-generator to create operational runbooks, with follow-up Q&A to refine both documents.

Multi-pass code review agent that combines security scanning, style review, and documentation generation in a single conversational session. Uses code-reviewer, security-auditor-pro, and doc-generator as tools.

Automated compliance audit agent that runs GDPR, license, and security checks on a codebase description or data flow. Synthesizes findings into a single risk-rated compliance report with remediation roadmap.

Full product discovery agent. Generates a complete PRD, extracts user stories, and breaks the MVP into an implementation task plan — all in one conversational session.

Audit AI systems and ML pipelines for regulatory compliance across EU AI Act, NIST AI RMF, and ISO 42001 frameworks. Identifies high-risk use cases, bias risks, transparency gaps, and human oversight deficiencies with a structured remediation plan.

Extract and summarize key terms from contracts, NDAs, SaaS agreements, and vendor agreements. Identifies obligations, liabilities, termination clauses, SLA commitments, IP ownership, and red-flag provisions with risk ratings.

Design GraphQL schemas from natural language data requirements. Produces type definitions, queries, mutations, subscriptions, input types, and pagination patterns following GraphQL best practices and Relay specification.

Generate descriptive, context-aware captions for images. Produces alt-text for accessibility, SEO-optimized descriptions, social media captions, and detailed scene analysis with object, action, and sentiment identification.

Professional-grade translation with deep context preservation. Handles technical documentation, marketing copy, legal text, and UI strings across 30+ languages. Preserves tone, formatting, brand voice, and domain-specific terminology.

Extract entities, relationships, and attributes from unstructured text to build a structured knowledge graph. Outputs nodes and edges in JSON-LD or Cypher format, suitable for Neo4j, GraphQL federation, or RAG enrichment pipelines.

Classify and intelligently route incoming emails by intent, urgency, department, and sentiment. Handles support tickets, sales inquiries, billing questions, abuse reports, and spam detection with configurable taxonomy and confidence scores.

Generate comprehensive inline documentation from source code. Produces JSDoc, TSDoc, Google-style Python docstrings, or Rust doc comments with parameter descriptions, return types, thrown exceptions, and usage examples.

Generate a complete product requirements document (PRD) from a raw idea. Includes user personas, jobs-to-be-done, feature list, technical architecture, success metrics, and Mistral AI integration recommendations.

Design complete system architectures from requirements. Produces component diagrams (Mermaid), API contracts, data models, technology choices with rationale, and scalability analysis.

Transform raw training examples into Mistral fine-tuning JSONL format. Validates message structure, enforces role alternation, and outputs ready-to-upload training data.

Generate the complete Mistral fine-tuning API call (Python and curl) from a training dataset description. Produces ready-to-run code for submitting a fine-tuning job to the Mistral API.

Comprehensive security audit with STRIDE threat modeling, CVSS scoring, compliance gap analysis (OWASP, SOC2, GDPR), and a prioritized remediation roadmap. Pro-tier depth beyond basic scanning.