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.
Craig M. Brown
BlindOracle Agent Trust answers one question before you work with or hire an AI agent: what is its record?
What works in ChatGPT today
Hiring, jobs and paymentThose happen over MCP, not in this plugin. Add https://api.craigmbrown.com/v1/mcp to Claude, Cursor, a ChatGPT Business or Enterprise workspace, or any MCP client. The first calls need no account; the steps are at https://craigmbrown.com/blindoracle/quickstart.html.
Proof of delegationWhen one agent hands a task to another, BlindOracle can record it as four signed steps: record, accept, deliver and close. Each step is a credential signed by did:web:craigmbrown.com that carries the hash of the step before it, and anyone can check a chain at https://api.craigmbrown.com/v1/delegations/ with only https://craigmbrown.com/.well-known/did.json. The prove-a-delegation skill explains the steps. Creating a chain is a paid action that happens over MCP at https://api.craigmbrown.com/v1/mcp, not in this plugin; the plugin itself stays read-only. A chain proves who paid and what each side presented. It does not prove either party's identity off BlindOracle, that a human was present, or that the work is good. Delegation tokens are secrets and should never be shared in public.
How it worksThe lookups are read-only. They share a daily limit across all ChatGPT users; when it is reached the tool says so and you can try again the next day. Every result carries a content credential signed by did:web:craigmbrown.com, and the result card shows that the signature is attached. It does not verify the signature in your browser; anyone can verify it against https://craigmbrown.com/.well-known/did.json.
What it does not doThe plugin never asks for wallet keys, seed phrases, card details or access keys. It does not read your chat history; a call sends only the tool arguments you provide.
Who it is forPeople and teams choosing agents to delegate work to, marketplace operators vetting counterparties, and developers building agent-to-agent workflows who want an independent record before the first transaction.
LimitsBlindOracle's record covers agents that have transacted with or been audited through BlindOracle. A missing record means "no record here", not "unsafe". Badges expire and are reissued; the expiry date is shown with each badge.
Fraud Blocker
Verify suspicious customer, lead, and signup email addresses directly in ChatGPT with Fraud Blocker.
Fraud Blocker Email Verification helps you investigate individual email addresses before deciding how to handle an order, signup, or lead.
Use Fraud Blocker when reviewing a suspicious ecommerce order, lead, registration, or customer email and you want additional information before taking action.
The plugin verifies individual email addresses. It does not automatically block orders, modify checkout flows, or take actions against users on your behalf.
DMARCwise
Inspect monitored domains, DMARC statistics and sending sources, SPF records and lookup chains, and DKIM selectors. Enable, update, or disable hosted DMARC using your authorized DMARCwise organization.
DMARC, SPF & DKIM Agent
Have I Been Squatted
Have I Been Squatted finds attacker infrastructure while it is being set up. Lookalike domains, mail servers, staging hosts and their certificates often appear before a phishing campaign launches. This plugin brings detection, investigation and response for them into ChatGPT and Codex.
Enrich and hunt, for any domain
Work your organization's detections
Background jobsScans and deeper analyses take longer than a single reply. They start a background job and return straight away; ask for the status later in the conversation to get ranked results. Jobs expire after 24 hours, and each organization runs three at a time.
ChangesTools that change your organization's data need an organization admin, and ChatGPT asks you to confirm each one. Takedown cases are reviewed by the Have I Been Squatted operations team before any report reaches a registrar or host. Sending a response changes your own security tools; the monitored domain itself is never blocked. Removing a monitored domain or deleting a rule cannot be undone.
Example prompts
Access requirements
Aikido Security
Scan the code you write or change for vulnerabilities, exposed secrets, and infrastructure-as-code (IaC) issues. Connect to Aikido Security to pull the matching findings from your feed
SOC 2, HIPAA & ISO readiness
Spot fraud & phishing signals
Evidence-first CVE triage
Agent Tools
Triage security vulnerabilities: by CVE id (how severe it is, whether it is being exploited, how likely exploitation is) or for a whole dependency list (which 3 packages to upgrade first, and to what version).
Ask "how bad is CVE-2021-44228?", "which of these CVEs should I patch first?", "what did CISA add to its exploited list this month?" or "which dependencies in this package-lock.json should I fix first?".
CVE Risk Check reads two free public sources. The National Vulnerability Database (NVD) gives the description, the CVSS score and severity, the publication date, a link to a patch or vendor advisory when one is tagged, and whether the CISA Known Exploited Vulnerabilities (KEV) catalog lists the CVE, with CISA's due date and required action. FIRST's EPSS gives the modelled probability that the CVE is exploited in the next 30 days.
Each CVE gets a priority: "exploited" when CISA lists it, "likely" when its EPSS score is 0.1 or more, otherwise "routine", or "unknown" when EPSS could not be read. A batch of up to 10 CVEs comes back sorted with the most urgent first.
Every answer names its sources and the date the NVD record and the EPSS score are from. NVD limits requests without a key, so records are cached for a few hours; a recent change to a CVE can take that long to appear, and when NVD asks this service to slow down the answer says so and when to retry.
For a dependency list, pass the text of a lockfile (package-lock.json, pnpm-lock.yaml, yarn.lock, poetry.lock, uv.lock, Cargo.lock, go.sum or go.mod), a package.json or a requirements.txt, or a list of packages with exact versions (npm, PyPI, Go, crates.io, Maven, RubyGems, NuGet or Packagist). Each package version is matched to its known advisories in OSV.dev, which collects the GitHub Advisory Database and other open databases. The advisories' CVE ids are then checked against the CISA exploited list and EPSS. The answer is the few packages to fix first, each with a priority (malicious, exploited, likely, routine or unknown), a one-line reason and the version to upgrade to, followed by a short list of the rest. A match is by exact name and version: it does not show that your code reaches the vulnerable code.
It does not scan your systems, read your source code, confirm that you are vulnerable, or tell you how to exploit anything. The CVE ids and dependency lists you send are not stored; the only lasting record is a daily count of calls per tool.
ArrowMem
Security and privacy tools for people running AI on their own hardware, plus answers about ArrowMem's privacy design. You can check whether your self-hosted model server is reachable from outside your machine, check whether it still matches the hardening guide, read the live-tested hardening procedure itself, and list the AI tools configured on your machine from config output you paste. You can also ask what ArrowMem is and who operates it, how it handles retention, encryption and where processing happens, what ArrowMem's run-it-on-your-own-hardware setup does today and what is planned, which government disability-claim packs run inside ArrowMem, what the open Data Dignity Standard (v0.5 draft for public comment) says, and verified facts about a local business, entered by hand from its own published sources with its consent, each with its source and date. A security alert relay can also send a push notice to you or your family. The check tools work only from diagnostic output you paste; they do not connect to your machine or network.
Cloudaware
Cloudaware Service Graph connects ChatGPT to a live inventory of the organization's infrastructure: every resource across AWS, Azure, Google Cloud, Oracle Cloud, Kubernetes, VMware and Nutanix, with its configuration, configuration history, cost, vulnerabilities, and the monitoring and security agents installed on it. Cloudaware discovers it through the providers' APIs and the systems already in place, keeps it current as resources change, and maps each resource to the application it serves. With the connector enabled, ChatGPT answers questions about inventory, configuration, changes, security coverage and cost across every cloud and data center from a single source. The connector is read-only and replaces a separate integration for every cloud and tool, so a single answer can draw on all of them.
Mailphantom
Mailphantom helps keep your real email address private with email aliases called phantoms. Email sent to a phantom is forwarded to your usual inbox.
The Mailphantom plugin lets you create and manage phantoms directly in ChatGPT. Ask for an alias for a new signup, find one you already use, or pause and resume forwarding—all through conversation.
Visit the Mailphantom website to learn more.
Check website tracking
AG Cyber Ltd
CIMTK helps cyber incident managers and response leaders manage active incidents. It provides guided workflows for executive briefings, public communications, incident-response advisory reviews, threat-intelligence requirements, supply-chain incidents, staff-message review and tabletop exercise design. This skills-only plugin works with information supplied by the user and does not connect to external systems or perform technical response actions.
nebty GmbH
Fake shops, phishing sites and fake profiles that copy your brand cost you customers, money and trust. nebty is the early warning system against brand impersonation: it scans the internet around the clock for lookalike domains, phishing websites, fake social media profiles, scam ads and malicious search results, scores every finding by risk and gets the dangerous ones taken down with registrars, hosting providers and platforms.
With this connector, your brand protection is one question away. Ask which lookalike domains pose the highest risk right now, which fake Instagram, Facebook or LinkedIn profiles and Meta ads were found, and how your takedowns are progressing, with their complete history. Review findings directly in the chat: mark profiles and ads as trusted or dismiss false positives, always within your team's existing permissions.
Connect your nebty account in seconds via secure OAuth. New scans and takedown requests run in the nebty dashboard, where nebty handles the removal for you.
Scan before you install.
Security reviews by VIDOC
Anomity
Anomity shows a security team every AI tool running on its managed devices: coding agents, MCP servers, skills and instruction files, browser AI use, personal versus corporate AI accounts, and the credentials those tools can reach. Sensors deployed through your device management report that inventory to the Anomity console, where policies turn it into findings. This plugin reads that data for your organization.
Brief leadership. Get a posture briefing built from live data: devices reporting, open and critical findings, personal AI accounts, unknown MCP servers, plaintext secrets found, and compliance scores.
Triage findings. Group open findings by the policy that raised them, rank them by severity and reach, and get the practical fix for each group.
Investigate a device. Name a laptop or a person and see the AI tools, MCP servers, findings and recent changes on that machine.
Review MCP servers. See which servers are official, community or unknown, how they run, what they can reach, and how many devices use them.
Work the allowlist. See what is waiting for approval across AI tools, AI sites, model providers, MCP servers, extensions, plugins and skills. Admins can approve the items they confirm, when an owner has turned that on in Anomity Settings.
Audit changes and compliance. Ask what changed in your AI tooling over the last 90 days, and where you stand on SOC 2, ISO 42001, the NIST AI RMF, the EU AI Act and the OWASP LLM Top 10.
You sign in with your Anomity console account, and you see what your console role allows, scoped to your organization. Secret values never leave the device and are never returned. Every tool reads data except one: approving allowlist items, which asks you to confirm the exact items first.
AgentGuards
AgentGuards checks the things an AI agent is about to trust or run.
Who it's for: developers using ChatGPT or Codex with untrusted inputs or shell access.
Requires an AgentGuards account (you sign in when you connect). Scan results also appear in your AgentGuards dashboard.
Limitations: the tools scan only what they are given, when they are called. They do not intercept anything automatically, and they never run, change or delete anything themselves.
Etugen
Etugen keeps track of malicious infrastructure on the internet, like command-and-control servers, malware hosts, phishing kits and attacker directories left open to the web. Connect your Etugen account and you can look all of this up right from the chat.
Give it an IP and it tells you which ports are open, which network and country it sits in, what malware runs there, its TLS and JARM fingerprints, how risky it looks and why, and which other hosts share the same certificate or fingerprint. You can also search by malware family, network, country or port, check who registered a domain, or see which ATT&CK techniques have been most active lately.
It's made for people who triage alerts, enrich indicators or write up investigations. Access is read-only and follows your Etugen plan. If we have no record of something, that means we haven't seen it, not that it's safe.
Detect VPNs, proxies and Tor
Phishing URL scans and pivots
Oneleet
Oneleet is a security and compliance platform that runs SOC 2, ISO 27001 and other compliance programs alongside penetration testing, attack surface monitoring, code security and device management. The Oneleet plugin connects ChatGPT and Codex to your Oneleet workspace so you can review and work on your security program in conversation.
What you can do:
How it works: you sign in with your Oneleet account and choose the workspace to connect. The plugin acts as you. It can only see and change what your role in that workspace allows, and every action is attributed to your account. Tools that create, change or delete data are marked as write actions, so you are asked to confirm before they run.
Requirements and limitations: you need an existing Oneleet account that is a member of a workspace with AI assistant access enabled. The plugin works in one workspace at a time; to switch, disconnect and sign in again. It cannot set a control's status directly (status comes from the control's checks and evidence), sign policies for other people, or reach workspaces you are not a member of.
