"Créer des systèmes de prompt pour les modèles de langage (LLM)" matching MCP servers:

A local MCP server that packages LLM evaluation gates as reusable CI/CD primitives, enabling AI agents to run datasets against models, score responses, and enforce quality thresholds.

A modular MCP server that extends GitHub Copilot's capabilities through intelligent context compression and dynamic model routing for long-lived coding sessions.

Um servidor MCP completo para integração com Sentry no Cursor, oferecendo 27 ferramentas para monitoramento de erros, performance e saúde de aplicações.

Exposes Registry Browser operations (embed, de-embed, publish, create packages) as MCP tools for AI agents and IDE integrations to manage Unity packages programmatically.

Turns Claude Code into an interactive OpenWiFi triage assistant, providing tools for searching Linux wireless patches, Jira tickets, running LLM-based triage, backporting, and building OpenWiFi.

Scan codebases for LLM API calls and estimate monthly costs. Compare costs between git refs to catch cost regressions during code review.

Combines GitHub releases from multiple repositories into a single product release note, with context sources for LLM synthesis.

Enables CI/CD failure triage for GitHub Actions by inspecting failed runs, classifying root causes with LLM assistance, retrieving relevant runbook guidance, estimating flakiness, and either preparing dry-run fix PRs or publishing structured diagnoses to Discord for human review.

Analyzes, plans, and benchmarks LLM workload migration from x86 to Arm64, generating measurable speedups using free GitHub Arm runners.

Agentic data quality MCP server — runs structured validation rules against warehouses (DuckDB, BigQuery, Athena, Databricks, Postgres), diagnoses failures with LLM root cause analysis, and proposes SQL remediations. Full audit trail of every AI decision.

MCP server for the Bitbucket Cloud REST API v2.0, providing 32 tools across 5 toolsets for repository management, pull requests, issues, pipelines, and more with safety controls and LLM-optimized output.

Enables LLM clients to access DevSecOps tooling such as CI/CD pipeline status, vulnerability triage, log search, and dependency scanning through MCP, turning AI copilots into security-aware engineering partners.

An MCP server that enables LLM coding agents to deploy web applications to AWS serverless infrastructure, supporting backend, frontend, and fullstack deployments.

Security scanner and MCP server that catches dangerous patterns in MCP servers and AI agent projects, such as leaked secrets, shell execution, and prompt-injection text. Runs as both a CLI and MCP server with CI-friendly severity gates.

Budget & cost control for AI agents: hard per-agent spend caps, rate limits, idempotency, and human-in-the-loop approval — enforced before each LLM call, not after the invoice. One hosted MCP endpoint (no proxy or self-hosting), settled via x402 (USDC on Base).

Enables Claude and other MCP clients to manage Pankow organizations, projects, test cases, and CI/runs, configure AI targets, and run LLM evaluations directly from an agent.

An MCP server for interactively reading, validating, previewing, and saving Renovate configurations with LLM assistance.

Gives an LLM agent a fixed, typed set of on-call tools over a self-hosted GitOps Kubernetes platform, reading metrics, logs, alerts and deploy history from Prometheus, Loki, Alertmanager and GitHub without any cluster write access. Rollbacks and config changes are proposed as one-shot ids that require explicit human confirmation before becoming pull requests against the Git repository, with both steps recorded in an audit log.

Enables intelligent Kubernetes deployments, automated documentation testing, organizational pattern management, and shared prompt libraries. Provides AI-powered recommendations based on cluster capabilities and automates DevOps workflows through conversational interfaces.

Agent trace diff and prompt regression detection — compares two agent runs to find exactly where reasoning diverged.