Un cycle Brainstorm → Spec validée → Plan à cases à cocher → TDD strict → Review de conformité, conçu pour les devs qui construisent des SaaS pilotés par IA (générateurs, agents, outils no-code) en HTML/CSS/JS vanilla ou Node/Express, avec intégrations multi-fournisseurs (Anthropic, OpenAI, Gemini, Mistral). Exemples d'usage : "ajoute une fonctionnalité X à mon générateur", "intègre un nouveau fournisseur IA", "refonds ce module"
Before you install a third-party skill, scan it. Skill Security Audit audits a skill package — code and SKILL.md prose alike — across 10 risk classes: plain-text instruction poisoning, hard-coded secrets, prompt injection including delegated sub-agent calls, destructive commands, data exfiltration, download-and-execute chains, over-broad permissions, and disclosure gaps. Judged under your agent's permissions. Reports findings with file and line evidence. Never runs the code it audits.
Hands-on prompt engineering and LLM evaluation. Built around a diagnose-then-ship loop: forensic transcript critique ▎ that names the exact failure mode, paired with a fixer that produces complete, deployment-ready rewrites — not ▎ vague advice.
Audit AI agents, Skills, MCP configurations, tool permissions, and coding-agent setups for prompt injection, excessive agency, credential exposure, data leakage, unsafe code execution, supply-chain risks, memory risks, missing approval controls, and runaway autonomy. Get a calibrated Security Score, Audit Confidence, evidence-backed findings, remediation steps, verification checks, and residual-risk analysis.
Turn messy notes, prompts, chat history, SOPs, and project docs into a reusable AI agent skill. Get a clean SKILL.md, reference-file plan, privacy review, and test prompts for Claude, Codex, Capafy, or local agents.
Turn prompts or workflows into testable, launch-ready AI Skill products.
Vibe Code Reviewer — Reviews AI-generated code for the mistakes that sink vibe-coded apps: exposed API keys, missing auth, SQL injection, client-trusted payments. Built for non-engineers — plain-language findings, severity ratings, copy-paste fixes. Catches the security holes behind the "45% of AI code is vulnerable" stat before you ship.
Turn pasted AI evaluation cases into an evidence-bounded triage brief that separates expected and observed behavior, hypotheses, and missing checks without claiming a root cause or fix.
Turn incomplete Jira tickets, JSM requests, PRDs, User Stories, and meeting notes into requirements that Engineering can build, QA can test, and Product can accept. Automatically identify requirement gaps, business rules, permissions, validations, UI/API behaviors, exceptions, and edge cases, then generate User Stories, Given/When/Then Acceptance Criteria, PM Checklists, QA Test Cases, risk analysis, and a Requirement Readiness Score based on the selected mode.