Creed logo

Creed

Freemium
Visiter

Creed est un profil de contexte personnel pour les agents IA, qui leur permet de répondre avec le même contexte dans tous vos outils au lieu de repartir de zéro. Disponible gratuitement en auto-hébergement ou via des offres hébergées pour particuliers et équipes.

Qu'est-ce que Creed ?

Creed is a personal context file for AI agents. It gives connected tools a shared profile that describes who you are, what you are working toward, how you like AI to respond, and the routines or constraints that should shape its answers.

The product is built around a simple loop: you create a concise Markdown profile, agents read it before they respond, and they can propose small updates when they learn something durable. That keeps the context current without forcing you to repeat yourself in every tool.

Que peut faire Creed ?

Structured context profile

Creed stores personal context as a structured Markdown profile with sections for identity, goals, work, preferences, routines, and optional sections like beliefs, constraints, people, health, and broader context.

MCP-based agent access

Connected agents read the profile before answering, so they can use the same background across tools instead of starting from zero in each conversation.

Proposal-based updates

When an agent learns something durable, it can propose a focused edit to the right section rather than rewriting the entire file.

Section quality scoring

Each section gets a quality score and signal review to help sharpen vague or thin writing over time.

Review and permission controls

Users can approve proposed changes, reject them, or allow trusted agents to edit directly on selected sections.

Hosted workspace options

Hosted plans add cross-device sync, backups, managed auth and storage, and an OpenRouter BYOK option for model spend.

Cas d'utilisation

“Personal AI context”

Set up a concise profile that captures your role, goals, preferences, and routines so agents can answer with the right context from the start.

“Cross-tool continuity”

Connect your everyday tools and let them read the same profile, reducing repeated explanations as you move between assistants.

“Reviewable context maintenance”

Use proposal-based edits to let AI capture durable preferences or routines without letting it rewrite the file unchecked.

“Shared team context”

Give a team one shared Company Creed so every member’s agents follow the same conventions, constraints, and working style.

“Hosted workspace management”

Keep context current on hosted plans with sync, backups, managed auth and storage, and optional BYOK if you want model spend on your own key.

Questions fréquentes

What is Creed used for?

Creed is a personal context profile: a concise Markdown file that captures who you are, what you are working toward, how you like AI to respond, and the routines and constraints it should respect. It is designed to be read by connected agents before they answer.

How does Creed stay up to date?

You can write the profile yourself through onboarding, then connected agents read it over MCP and propose small updates when they learn something durable. You approve the good changes, or grant direct-edit access to trusted agents on specific sections.

Which tools does Creed work with?

Creed connects to agents such as Claude Code, Codex, Cursor, and ChatGPT over MCP, and it integrates with GitHub for version control. The docs also mention that support for Notion and Obsidian is coming for editing and storage.

Is Creed only for individuals?

Yes. The Company plan adds one shared Company Creed, member roles, section permissions, and an activity view. Members can read the shared file through their own agents, while owners and admins control billing and content.

Does Creed require its own AI spend?

Hosted plans include usage credits and support BYOK, which means you can use your own OpenRouter key so model spend runs on your account. If you stay within the included allowance, the hosted plan covers the usage; otherwise you can top up or use your own key.

Informations rapides

Category
AI context management
Primary users
Individuals and teams using connected AI agents
Product format
Plain Markdown profile with structured sections
Connection method
MCP
Source domain
creed.md
Deployment
Self-hosted open source build and hosted plans

Analyse du trafic de Creed

Les données de trafic sont fournies à titre indicatif uniquement.

Évaluation du domaine
2

Alternatives à Creed

CellCog logo

CellCog

cellcog.ai

CellCog est une plateforme généraliste d’employés IA qui permet de recruter des agents IA par rôle, chacun disposant de sa propre boîte de réception, de son tableau de tâches et de sa mémoire. Elle s’adresse aux équipes et aux fondateurs qui veulent confier à l’IA des tâches de connaissance continues et la coordination avec d’autres employés.

Kortyx logo

Kortyx

kortyx.co

Kortyx est un assistant IA proactif qui apprend votre façon de travailler et capte le contexte de votre activité sur ordinateur et dans le navigateur. Avec votre accord, il peut suggérer et effectuer des tâches en s’appuyant sur votre historique, vos brouillons et vos pages pour vous aider à terminer le travail courant ou inachevé.

Google Antigravity logo

Google Antigravity

antigravity.google

Google Antigravity est une plateforme de développement agentique pour les agents autonomes dans les apps, CLI, IDE et SDK, avec téléchargement Windows, offre individuelle gratuite et accès supérieur pour les utilisateurs Pro, Ultra et les organisations Google Cloud.

Agno logo

Agno

agno.com

Agno est un framework Python open source pour créer des agents IA et des workflows multi-agents, avec AgentOS pour les exécuter et les gérer comme des services API en production dans votre cloud.

Context logo

Context

context.ai

Context est une plateforme d’agents IA d’entreprise pour créer, déployer et améliorer des agents sur l’infrastructure client, avec espace de travail, runtime, contexte, évaluation, connecteurs et contrôle d’accès basé sur l’IdP.

Spell logo

Spell

spell.so

Spell est une plateforme web d’agents IA pour déléguer des tâches de connaissance à des agents autonomes, avec plugins, tâches parallèles et prise en charge de GPT3.5/GPT4.