Software Development Skills

Software architecture: design, trade-off analysis, and decision documentation. Use this skill whenever the user asks to design an architecture, compare technical options, make an architecture decision, analyze trade-offs, write an ADR, model a system, choose between patterns or stacks, plan a migration, or any question involving a structural decision about software design. Also trigger when the user says: 'architecture', 'how to structure', 'which pattern', 'monolith vs microservices', 'ADR', 'bounded context', 'how to migrate', 'which stack', 'technical trade-off', 'architecture decision', 'conceptualize', 'design the system', 'system design', even if the word architecture is not explicitly mentioned.

Transform a DOCUMENTOR research report into a client-ready deliverable. Trigger when the user asks to: "brief", "reformat for the client", "turn it into a deliverable", "write up the brief", "client format", "write a brief", "make it client-ready", "prepare the deliverable", "creative brief", "brief créatif", "design brief", "brief design", or any phrasing implying transformation of research into a polished output for an external audience. Do not use for internal research — that's DOCUMENTOR's job.

Builds, audits, and formalizes enforceable functional specification documents (cahier des charges — CdC, CdCF, CCTP) that comply with NF EN 16271 and ISO/IEC/IEEE 29148:2018. Three modes: Generation (turns a brief into a complete, ready-to-sign CdC), Audit (scores an existing CdC against the 14-criteria grid plus quantitative KPIs), and Legal (a contractual-enforceability checklist). Produces a branded PDF deliverable via a PDF generator skill. Use this skill whenever the user asks to: draft a cahier des charges, write a CdC, create a CdCF, formalize a client's needs, prepare a CCTP, audit a cahier des charges, score a CdC, check the quality of a cahier des charges, make a CdC enforceable, prepare a contractual cahier des charges, "write a cahier des charges", "generate a CdC", "audit this cahier des charges", "turn this brief into a cahier des charges", "prepare the cahier des charges for this client", "make me a functional cahier des charges", "draft a cahier des charges for my vendor", "check that this CdC holds up legally". Do NOT use it for purely technical SRS documents aimed at an internal dev team (that is srs-writer) or for marketing client briefs (that is brief-writer). This skill covers the formalized expression of needs between a client and a provider, with contractual value.

A CEO advisor who challenges product decisions, prioritization, roadmap features, pivots, and strategic trade-offs. Use this skill whenever the user is hesitating about building a feature, adding a product capability, pivoting, killing a project, choosing between two directions, or when they say \"I'd like to add\", \"I'm thinking about\", \"should I\", \"next step\", \"roadmap\", \"prioritization\", \"feature request\", \"product idea\", \"what do you think of this idea\", \"we could\", \"it would be cool if\". Especially suited when the user is looking for a critical perspective rather than validation — this skill exists to push the decision, not to rubber-stamp it. Trigger it proactively even if the user doesn't explicitly ask to be challenged — as soon as they describe a product decision in progress, activate. Do NOT use it for pure technical architecture questions (that's senior-dev) or for documented market research (that's DOCUMENTOR). This skill is for the decision itself, not for the analysis that precedes it.

A lightweight synthetic CAB (Change Advisory Board) for a solo developer working with AI agents. Assesses the risk of a change, produces a structured recommendation (deploy/hold/investigate), and orchestrates post-deployment verification. Use this skill whenever the user wants to: deploy to production, ship a major version upgrade, merge a large PR, push infrastructure changes, modify a production database, upgrade a critical dependency, or take any action whose rollback would be costly. Also when the user says "can I deploy this?", "is it safe to push?", "review before deploy", "change advisory", "risk check", "pre-deploy check", "is this production ready?", or hesitates before a deployment.

CLAUDE.md file generation and optimization for Claude Code projects. Capabilities: initialize project instructions, analyze codebase context, optimize existing CLAUDE.md, apply Anthropic best practices, reduce token usage, improve effectiveness. Actions: init, create, optimize, enhance CLAUDE.md files. Keywords: CLAUDE.md, project instructions, Claude Code setup, project context, codebase analysis, Anthropic best practices, token optimization, project configuration, AI instructions, coding guidelines, project rules, workspace setup. Use when: initializing CLAUDE.md for new projects, optimizing existing project instructions, setting up Claude Code for a codebase, improving AI coding guidelines.

Generate ultra-detailed, copy-paste-ready prompts that produce stunning landing pages, hero sections, and web interfaces when fed into AI coding tools (Bolt, Lovable, Claude Code, Cursor, v0). Use this skill whenever the user asks to: generate a design prompt, create a prompt for a landing page, write a prompt for Bolt/Lovable/v0, 'make me a prompt for...', 'a prompt for a landing page', 'generate a hero section prompt', 'create a design prompt', or any phrasing implying the creation of a detailed frontend prompt meant to be used in an AI code generation tool. Also trigger when the user says 'MotionSites style', 'a prompt like MotionSites', or wants a reusable prompt template for web design. Do NOT trigger when the user wants you to directly build/code the interface — that's the frontend-design skill.

Docker and Dokploy expert for deploying complex infrastructure on self-hosted servers. Use this skill whenever the user mentions: Docker, Dockerfile, docker-compose, docker compose, container, Dokploy, deployment, deploy, Traefik, reverse proxy, SSL, Let's Encrypt, certificate, domain, multi-service, microservices, Docker Swarm, cluster, scaling, replicas, health check, volume, bind mount, Docker networking, bridge, overlay, Docker image, multi-stage build, registry, Docker CI/CD, deploy webhook, zero-downtime, rolling update, rollback, port mapping, expose, docker prune, Docker logs, container monitoring, resource limit, OOM, restart policy, sidecar, init container, docker secret, or any question involving Docker in development or production, or deploying applications and databases on Dokploy. Also trigger when the user writes a Dockerfile or a docker-compose.yml, configures Traefik, debugs a "Bad Gateway", plans a multi-container architecture, or prepares a deployment to a VPS. Do NOT use for purely PostgreSQL questions with no Docker context (that's postgres-expert) or for abstract software architecture unrelated to deployment (that's architecture-conceptor).

Use this skill for any web research, monitoring, source synthesis, information comparison, or production of documented and sourced deliverables. Trigger automatically when the user asks to: "research", "find", "search", "synthesize", "compare sources", "what is", "what's the state of", "monitor", "document", "analyze the market", "benchmark", or any phrasing implying factual research with citations. Applicable to technical, regulatory, scientific, and business topics. Do not use for pure content generation without factual grounding.

Rigorous product audit of an already-built feature to detect UX friction, functional gaps, and user risk, then produce actionable, prioritized, and justified recommendations (never generic). Use this skill whenever the user asks to audit a feature, polish a feature, finalize a feature, challenge a feature before ship, detect friction, identify functional gaps, or says "audit this feature", "what's missing", "where's the friction", "challenge this feature", "pre-ship review", "feature review", "put it through its paces", "what's wrong with this", "product review", "UX audit", "product audit", "functional audit", "what's missing", "what could go wrong", "friction audit", "critique this feature" (in the product sense, not visual). Also trigger when the user wants a second critical read before delivery, wants to validate that a feature is ready, or wants to identify the blind spots in a spec. Do NOT use for purely visual critique (→ critique), technical code audit (→ senior-dev, audit), security audit (→ pentest-audit), performance audit (→ perf-analyzer), or design from scratch (→ frontend-design).

Generate a justified, research-backed business model for an existing web or mobile project. Use when the user asks to define, challenge, compare, or create a business model, monetization model, pricing strategy, Business Model Canvas, go-to-market economics, or investor/client-ready business model report for a digital product, SaaS, marketplace, mobile app, AI app, ecommerce, subscription app, community, platform, or service. The skill guides project analysis, targeted user questions, web research, evidence scoring, country/sector differences, model selection, justification of tradeoffs, and creation of a PDF report authored by the AI agent.

Implement, configure, and maintain the LLM Lab for testing skills across multiple LLMs, comparing their results, and building autonomous pipelines. Use when the user mentions LLM lab, Ollama, LiteLLM, Langfuse, DeepEval, local inference, local model, LLM gateway, routing, benchmark, LLM evaluation, LangGraph, skill portability, model comparison, adding a model, lab configuration, lab docker-compose, observability, or LLM tracing. Trigger it as well for any need to set up, maintain, or experiment across multiple models.

Mobile-responsive rework of existing React components. Analyzes a desktop-first component and turns it into a mobile-optimized one by applying standards for touch ergonomics, WCAG accessibility, animation performance, and fluid layout. Use this skill whenever the user asks to: make a component responsive, adapt a page for mobile, fix the responsive behavior, optimize for smartphone, rework the mobile experience, "it doesn't work on mobile", "it's broken on small screens", adapt a form for mobile, or any request that involves adapting an existing React interface for touch smartphone screens.

In-depth security audit and penetration testing of web, API, and mobile applications. Static source-code analysis, dynamic testing, configuration checks, and generation of structured reports with severity ratings and recommendations. Use this skill whenever the user mentions: pentest, security audit, vulnerability, OWASP, SQL injection, XSS, CSRF, SSRF, IDOR, broken access control, security review, security headers, CSP, CORS, TLS, exposed secrets, fuzzing, security scan, 'is this secure', 'check the security', 'find the flaws', 'audit this code', 'penetration test', 'security check', or any question involving the security of an application, an API, or application infrastructure. Also trigger when the user shares code and asks whether there are potential security issues, even implicitly. Do NOT use for pure architecture questions (that's architecture-conceptor) or for code review without a security angle (that's senior-dev).

Use when synthesizing perf findings into evidence-backed recommendations and decisions.

PostgreSQL expert for query optimization, indexing, extensions, and database…

Product advisor for making an app as relevant as possible. Helps structure a project's foundations, find ideas for innovative features, enrich an existing idea, and validate business/technical relevance. Use this skill whenever the user asks: "how to improve my app", "feature idea", "which feature to build", "enrich this idea", "is this feature worth it", "how to make my product better", "product advisor", "brainstorm feature", "innovative idea", "what to build next", "how to monetize", "how to validate this idea", "feature discovery", "product opportunity", "improve the experience", "which direction to take", "pivot or persevere".

Guides the launch of a project — whether an internal product or a client engagement — through a 5-phase workflow (Framing / Discovery / MVP / PMF / Growth) that draws on the playbooks of Altman, Graham, Cagan, Ries, the Mom Test, Working Backwards, the Design Sprint, and Sean Ellis's 40% rule. Use this skill as soon as the user says: "launch a project", "kickoff", "how do I get started", "I have an idea", "validate an idea", "MVP", "PMF", "product-market fit", "discovery", "should I build X", "start a project", "launch", "go-to-market", "launch a product", "onboard a client", "first project", "from zero to one", "scope a project", "frame a project", "where do I start with my project", or any phrasing that involves the structured kickoff of a new project/product/feature/client engagement. Includes hard gates (≥5 Mom Test interviews before Phase 2, the 40% rule measured before declaring PMF) and warnings about the classic anti-patterns (schlep blindness, vanity metrics, founder mode break, premature hiring). Do NOT confuse with ceo-challenger (challenging existing decisions), architecture-conceptor (technical design), apply (backlog execution), generate-business-model (business model), senior-dev (implementation decisions). Covers the what / why / how-to-validate, not the how-to-code.

Production-ready technical decisions like a senior engineering lead. Use for: architecture choices, implementation planning, trade-off analysis, incident remediation, rollout strategy, code review, decision validation. Triggers on 'how should we', 'best approach', 'choose between', 'plan the rollout', 'what are the options', or French equivalents. Always use for high-stakes decisions impacting reliability, scalability, or maintainability. Skip for simple how-to, one-off debugging, or trivial edits.

Analyze a complete repository and create a SonarQube configuration tailored to the project type. Use this skill whenever the user asks to configure SonarQube, SonarCloud, sonar-project.properties, quality profiles, quality gates, Clean as You Code, coverage, duplication, security hotspots, custom rules, OWASP/CWE/MISRA/CERT/ISO 25010 standards, or wants to make a SonarQube analysis genuinely fit a repo. Even if the user only says 'wire up Sonar', 'set up the quality gate', 'configure static analysis' or 'make a clean sonar config', this skill applies.

Create, audit, and maintain engineering standards for a software project. Use this skill whenever the user asks to: create a standard, write a norm, define code rules, establish a convention, write a review standard, or a standard for CI/CD, security, monitoring, documentation, or any technical standard. Also trigger when the user says: "we should standardize", "create a standard for", "define the rules for", "norm for", "convention for", "quality standard", "code standard", "security standard", "PR rules", "standard process", "audit this standard", "update this standard", "review this standard", or any request involving the creation, audit, or maintenance of an engineering standard. Do NOT use for one-off architecture decisions (that is architecture-conceptor) or for pure research (that is DOCUMENTOR).

Expert in Linux and Windows systems infrastructure for administration, monitoring, security, and automation. Use this skill whenever the user mentions: Linux, Windows Server, sysadmin, system administration, systemd, services, crontab, scheduled tasks, GPO, Group Policy, Active Directory, AD, LDAP, firewall, iptables, nftables, Windows Firewall, UFW, SELinux, AppArmor, monitoring, Prometheus, Grafana, Zabbix, Nagios, journalctl, Event Viewer, performance, latency, CPU, RAM, disk I/O, network, bandwidth, NIC bonding, teaming, VLAN, DNS, DHCP, NTP, SSH, WinRM, PowerShell remoting, RDP, software installation, package manager, apt, yum, dnf, pacman, winget, chocolatey, scoop, MSI, GPO deployment, WSUS, patch management, automation, Ansible, bash script, PowerShell script, CI/CD pipeline, rsync, robocopy, scp, sftp, file transfer, ETL, data pipeline, cron, Task Scheduler, systemd timer, system role, RBAC, sudo, sudoers, permissions, ACL, NTFS permissions, certificates, TLS, PKI, hardening, CIS benchmark, security audit, compliance, backup, restore, disaster recovery, high availability, clustering, load balancing, or any question about systems infrastructure, server security, or automation. Finance/low-latency context: also trigger when the user talks about real-time data transfer, network latency, kernel tuning, TCP tuning, DPDK, SR-IOV, CPU pinning, NUMA, huge pages, real-time kernel, market data feed, FIX protocol infrastructure, colocation, or any system performance constraint related to finance. Do NOT use for: Docker/containers without a systems context (docker-dokploy-expert), pure PostgreSQL (postgres-expert), abstract software architecture (architecture-conceptor), or web application security (pentest-audit).

Break down and explain complex technical concepts using everyday analogies, diagrams, and structured mini-lessons. Aimed at L3 developers (1-3 years of experience). Use this skill whenever the user asks: 'explain to me', 'what is', 'break it down', 'how does it work', 'I don't get it', 'explain like I'm', 'ELI5', 'simplify', 'help me understand', 'why do we use', 'what's the difference between', 'what's it for', or any request that involves an educational explanation of a technical, architectural, or computing concept. Also trigger when the user seems confused about a concept or asks for a comparison between technologies/patterns. Do NOT use for factual research (that's DOCUMENTOR) or for architecture decisions (that's architecture-conceptor).