Classement des plugins
Consultez l’ordre du répertoire public des plugins de ChatGPT par date de collecte. Popular est une sélection éditoriale. Le répertoire ne publie aucun chiffre d’installation ou d’utilisation.
Resolve AI, Inc.
Connect ChatGPT and Codex to Resolve AI. Start and steer investigations, pull live production context, and turn root-cause findings into pull requests.The Resolve plugin connects ChatGPT and Codex to Resolve AI over MCP. Resolve's agents join your on-call rotations to triage and investigate alerts and work alongside engineers to reach root cause faster; this plugin brings that surface directly into your coding agent so you can act on it where you already code.It installs skills for listing and filtering alerts, reading and summarizing investigations, starting new root-cause analyses, asking follow-up questions, steering an active investigation with context you have locally, pulling an area's production profile before a risky change, and translating findings into local code changes and a PR.
Code Happy, LLC
Create images, website screenshots, and PDFs directly in ChatGPT with HTML/CSS to Image. Describe what you want, and ChatGPT writes the HTML and CSS and renders the result. ChatGPT can also create and manage reusable templates for you, update their designs, and render them with new content.
Use it to:
張聖杰
HOME AI Node lets ChatGPT inspect the health, runtime state, and tracked tasks of a HOME AI node. get_node_status reports whether a node is online and its repository state, get_runtime_state shows the runtime mode and MCP protocol, and get_task_status queries task progress. In this phase the server is backed by a deterministic synthetic review fixture (node review-demo-node) so reviewers can exercise every tool without any private network access; connection to a real HOME AI node is a later phase behind an authorized-user boundary (OAuth). All operations are strictly read-only: nothing on this server can write, execute, or reach any private network.
Knock
Knock is agent-first customer engagement infrastructure for your product, marketing, and transactional messaging. One platform to manage all your messaging across email, SMS, push, chat, and in-app.
Inspect workflows and manage notifications, directly in ChatGPT and Codex.
Suger Inc.
Suger helps marketplace teams work with cloud marketplace, CRM, co-sell, integration, billing, and knowledge data in ChatGPT. Users can look up offers, buyers, entitlements, referrals, integrations, workflow runs, CRM records, and support knowledge, and perform enabled actions through their connected Suger organization.
Uizze
Uizze gives coding agents a free anti-ui-slop Skill and optional full MCP access to live UI references, license-clear materials, and rendered safety checks. Use the skill to inspect the product and design system, write a design contract, cover required states, and run a bounded finish gate. The MCP answers one concrete unresolved UI question with strong full-screen references, finds license-clear materials for a concrete role, or optionally checks rendered UI for observable breakage, accessibility, privacy, and task-blocking composition. The free preview MCP exposes a deterministic source check only.
HOIPOI
A connector that lets AI agents search the official knowledge base of TALON, a low-code platform for building business applications. It searches thousands of pages of official manuals, support Q&A, a database of 900+ real-world function configuration examples, and specification analyses, and returns answers with source URLs as evidence. Typical uses include feature design consultation, configuration value lookups, error-cause investigation, browsing similar real-world examples, and automated validation of generated settings.
A TALON maintenance contract and a TALON support site (Zendesk) account are required. Users authenticate by signing in to the support site when connecting. All tools are read-only, and users' conversation content, code, and attachments are never sent to our servers (details: https://mcp.talon.jp/privacy).
FOR IO LABS INC.
Create fully working business software and web apps from a prompt. With Craftian you can create CRMs, task managers, inventory systems, expense trackers, time trackers, job portals, on-boarding portals, data forms etc.
Craftian creates production-ready web apps with built-in database, authentication, role-based access control, file storage, and automatic cloud deployment.
No coding experience required. Just describe what you want, and keep refining it.
INSPIREUI COMPANY LIMITED
Connect to the FluxBuilder MCP server to manage native Flutter app projects for WooCommerce, Shopify, and WordPress. Easily create apps, redesign screens, update settings, search store catalogs, capture mobile previews, and run test or release builds for the App Store, Google Play, and web—all without writing code.
Sleuth Inc.
Sleuth Skills connects ChatGPT to your personal skills.new vault — a private collection of skills, rules, agents, and slash commands you've authored or saved. The app lets ChatGPT discover what's in your vault and load the full content of a specific skill on demand, so conversations can follow your own playbooks instead of relying on generic knowledge. It is read-only: it never modifies, deletes, or shares your vault content.
BEIKE YAN
Creates evidence-based dependency upgrade preparation plans from supplied release, usage, testing, and rollback materials. It does not modify repositories, execute upgrades, run tests, commit changes, or create pull requests.
EzzyBills
Access the authoritative EzzyBills developer specification, REST API contracts, authentication models, and integration guidance for building EzzyBills applications such as approval dashboards, email response workflows, and purchase order tools.
Control Plane Corporation
Deploy and run cloud apps
Control Plane runs containerized apps across AWS, GCP, Azure, Oracle Cloud, and your own hardware as one platform. Build a GitHub or a GitLab repo into a container and deploy it to a live HTTPS URL, give it a custom domain, attach Postgres or Redis, and let it autoscale. Diagnose failures from logs, metrics, and events, and find the over-provisioned workloads inflating your cloud bill.
Capture customer issue context
BEIKE YAN
Turn supplied test output into a clear failure list, evidence-based classifications, and an investigation order. The plugin works only from the information you provide and does not access code, CI systems, repositories, or external services.
BitsWeave
Bitsweave is a user-owned context layer for AI tools. It stores the structured work of your organization, connects it via a context graph and exposes it to Claude over MCP.
Sign in with OAuth. ChatGPT can search and hydrate your org context so you do not repeat who you are, what the project is, or what you already decided.
What you can do:• Look up a specific fact about your company, a past decision, or a current task.• Ask a synthesis question across decisions, notes, and related records.• Create or update a Bitsweave record from the conversation when you want the result stored.
Connect GitHub, email, and other tools through Bitsweave brokers when you enable them. Claude calls Bitsweave. Bitsweave uses the permissions you granted. The agent does not hold the upstream credentials.
Sign-in is required. The connector URL is https://bitsweave.com/api/mcp.
Setup: add Bitsweave, complete the OAuth sign-in, then ask ChatGPT to search your context or create a record. Public docs: https://bitsweave.com/docs/use/onboarding/connect-your-agents
Prefect
Inspect Prefect workflows
Connect Prefect Cloud to inspect workflows, diagnose failed, late, or stuck runs, review workspace and worker health, and search current Prefect documentation and release notes. The hosted tools are read-only.
Redis
Design, implement, review, and troubleshoot Redis applications with focused guidance for data structures, connections, Redis Search, semantic caching, clustering, security, observability, and Iris agent memory.
Resolve website feedback
NeuronSearchLab
Inspect and manage your NeuronSearchLab recommendation workspace from ChatGPT. Search catalogue items, understand rankings, configure contexts, pipelines, rules and segments, run experiments and model training, and review performance analytics.
CrowdView, Inc. dba Extend
Turn documents into data
Parse, extract, classify, and split documents from chat: pull fields, tables, and line items out of PDFs, images, Office files, and spreadsheets, identify document types, break bundles into separate documents, and fill PDF forms. Build and test reusable extractors, classifiers, and workflows on real documents, then run them at scale in your Extend workspace.
Relewise ApS
Build, review, and troubleshoot Relewise Search, Recommendations, behavioral tracking, and data integrations with current API and SDK guidance for C#, TypeScript/JavaScript, PHP, and Java.
HORIZON SHIELD
This plugin exposes a public register of measured conduct for MCP servers.
Five conditions are measured on a schedule: the server speaks MCP, it publishes an A2A agent card, it declares who pays it, identical input returns identical output, and the verdict itself can be recomputed by a third party. A condition that could not be measured is recorded as not measured. It is never counted as a pass, and that rule is not relaxed for the operator of the register.
The register does not rank servers. It does not recommend one server over another. It returns what was measured, when it was measured, and the SHA-256 of the record so that the reading can be checked by someone who does not trust the party that issued it.
An endpoint that is absent from the register has simply never been measured here. Absence is reported as absence and is not a negative verdict.
Determinism is not measured unless the owner of the checked server has recorded consent, because measuring it requires executing a tool on someone else's server.
Free, read only, no account and no key. The register is a build artefact of a scheduled script with no editorial step, and it is archived with a DOI so a citation survives the repository moving.
OpenAI
WebMCP helps inspect a site's real journeys, propose a small task-completing tool surface, implement approved tools through existing trusted paths, and verify the result end to end.
