MentiSphere
mentisphere_wiki
https://mentisphere.wiki/wiki/Main_Page
MediaWiki 1.43.8
first-letter
Media
Special
Talk
User
User talk
MentiSphere
MentiSphere talk
File
File talk
MediaWiki
MediaWiki talk
Template
Template talk
Help
Help talk
Category
Category talk
Form
Form talk
Module
Module talk
Agent
Agent talk
Knowledge
Knowledge talk
Skill
Skill talk
Orchestrator
Orchestrator talk
Main Page
0
1
274
3
2026-04-25T14:28:59Z
Admin
1
Fix DPL namespace separator (comma -> pipe)
274
wikitext
text/x-wiki
<div class="ms-hero">
<div class="ms-hero-title"><span id="ms-hero-logo"></span>MentiSphere</div>
<div class="ms-hero-tagline">Where domain experts shape AI knowledge</div>
<div class="ms-hero-subtitle">The collectively-built AI intelligence platform</div>
</div>
<div class="ms-chat-input-wrapper" id="ms-home-chat"></div>
<div class="ms-action-cards">
<div class="ms-action-card">
<div class="ms-action-card-icon">🧠</div>
<div class="ms-action-card-title">Agent</div>
'''[[Special:FormStart/AgentPage|Create an Agent]]'''
<div class="ms-action-card-desc">Share your expertise as an AI agent</div>
</div>
<div class="ms-action-card">
<div class="ms-action-card-icon">📖</div>
<div class="ms-action-card-title">Knowledge</div>
'''[[Special:FormStart/KnowledgePage|Add Knowledge]]'''
<div class="ms-action-card-desc">Contribute reference material for agents</div>
</div>
<div class="ms-action-card">
<div class="ms-action-card-icon">⚙️</div>
<div class="ms-action-card-title">Skill</div>
'''[[Special:FormStart/SkillPage|Create a Skill]]'''
<div class="ms-action-card-desc">Define executable capabilities</div>
</div>
</div>
== Browse Agents ==
{{#dpl:
| namespace = Agent
| ordermethod = lastedit
| order = descending
| count = 10
| format = ,\n* [[%PAGE%|%TITLE%]],,
| noresultsheader = ''No agents yet. Be the first to [[Special:FormStart/AgentPage|create one]]!''
}}
[{{fullurl:Special:AllPages|namespace=3000}} View all agents →]
== Browse Knowledge ==
{{#dpl:
| namespace = Knowledge
| ordermethod = lastedit
| order = descending
| count = 10
| format = ,\n* [[%PAGE%|%TITLE%]],,
| noresultsheader = ''No knowledge pages yet. [[Special:FormStart/KnowledgePage|Add one!]]''
}}
[{{fullurl:Special:AllPages|namespace=3010}} View all knowledge →]
== Browse Skills ==
{{#dpl:
| namespace = Skill
| ordermethod = lastedit
| order = descending
| count = 10
| format = ,\n* [[%PAGE%|%TITLE%]],,
| noresultsheader = ''No skill pages yet. [[Special:FormStart/SkillPage|Create one!]]''
}}
[{{fullurl:Special:AllPages|namespace=3020}} View all skills →]
== Recent Activity ==
{{#dpl:
| namespace = 3000|3010|3020
| ordermethod = lastedit
| order = descending
| count = 10
| format = ,\n* [[%PAGE%|%TITLE%]],,
| noresultsheader = ''No recent activity.''
}}
== How It Works ==
No single AI lab can capture the full breadth of human expertise. A nurse in Naples, a logistics engineer in Singapore, and a farmer in Iowa each hold specialized knowledge that general-purpose AI lacks.
MentiSphere gives them a way to embed that knowledge directly into AI behavior — without writing code.
<div class="ms-steps">
<div class="ms-step">
<div class="ms-step-number">🧠</div>
<div class="ms-step-title">Agents</div>
<div class="ms-step-desc">Community-authored AI personas. Each agent is a wiki page containing a system prompt — written, refined, and voted on by domain experts. When you chat, the agent reasons using that collective expertise.</div>
</div>
<div class="ms-step">
<div class="ms-step-number">📖</div>
<div class="ms-step-title">Knowledge</div>
<div class="ms-step-desc">Reference material that agents can access. Facts, procedures, domain data — contributed by experts, retrieved automatically when relevant to your question. Think of it as a textbook the AI can read.</div>
</div>
<div class="ms-step">
<div class="ms-step-number">⚙️</div>
<div class="ms-step-title">Skills</div>
<div class="ms-step-desc">Executable capabilities that teach AI how to do specific things — format reports, structure assessments, follow protocols. Agents use skills on demand, combining reasoning with action.</div>
</div>
</div>
Wikipedia articles teach humans. Our pages teach AI. Same social dynamics, same governance, same tools — different artifact type. The unit of contribution is natural language. The contribution is executable.
5y4scklzf42rbdduixv3udt269ohej5