verdigraph
Verifiable cognition: any agent file to a deterministic content-addressed brain_id.
使うべきか
品質と安全性
検出事項（1）
- LOWTool 'describe_agent' description lacks action verbdescribe_agent 内
ツール定義とプロトコルへの準拠に関する自動分析に基づいています。
コンテキストコスト
これは、サーバーのツールがモデルのコンテキストに読み込まれるたびに消費されるおおよそのトークン数です。数が多いほど、ほかのタスクに使える注意が減ります。
インストール
ワンクリックインストール
これを `claude_desktop_config.json` ファイルに追加してください:
{
"mcpServers": {
"verdigraph": {
"url": "https://mcp.viridisconservation.com/verdigraph/mcp"
}
}
}リモートエンドポイント
https://mcp.viridisconservation.com/verdigraph/mcpstreamable-httpできること
ツール一覧
ツール（4）
🟢build_brain(content, format, include_document, request_id)
Compile an agent file into a deterministic, content-addressed brain. content = the file as a string; format = verdigraph_genome | claude_project_export | openai_assistant | prompt_list | auto. Returns brain_id, content_hash, node/edge counts, the 9-invariant firing report, and provenance. Identical bytes always produce the identical brain_id (VB1). include_document=true returns the full graph.
入力スキーマ
{
"type": "object",
"properties": {
"content": {
"title": "Content",
"type": "string"
},
"format": {
"default": "auto",
"title": "Format",
"type": "string"
},
"include_document": {
"default": false,
"title": "Include Document",
"type": "boolean"
},
"request_id": {
"anyOf": [
{
"type": "string"
},
{
"type": "null"
}
],
"default": null,
"title": "Request Id"
}
},
"required": [
"content"
],
"title": "build_brainArguments"
}出力スキーマ
{
"type": "object",
"properties": {
"result": {
"title": "Result",
"type": "string"
}
},
"required": [
"result"
],
"title": "build_brainOutput"
}🟢verify_brain(content, brain_id, content_hash, format)
Machine-check a cognition claim: recompute the brain from the submitted content and compare against the claimed brain_id and/or content_hash. valid=true iff every claimed identifier matches the deterministic recomputation (VB3).
入力スキーマ
{
"type": "object",
"properties": {
"content": {
"title": "Content",
"type": "string"
},
"brain_id": {
"default": "",
"title": "Brain Id",
"type": "string"
},
"content_hash": {
"default": "",
"title": "Content Hash",
"type": "string"
},
"format": {
"default": "auto",
"title": "Format",
"type": "string"
}
},
"required": [
"content"
],
"title": "verify_brainArguments"
}出力スキーマ
{
"type": "object",
"properties": {
"result": {
"title": "Result",
"type": "string"
}
},
"required": [
"result"
],
"title": "verify_brainOutput"
}⚪detect_format(content)
Detect which supported agent-file format the content is (verdigraph_genome, claude_project_export, openai_assistant, prompt_list) before building.
入力スキーマ
{
"type": "object",
"properties": {
"content": {
"title": "Content",
"type": "string"
}
},
"required": [
"content"
],
"title": "detect_formatArguments"
}出力スキーマ
{
"type": "object",
"properties": {
"result": {
"title": "Result",
"type": "string"
}
},
"required": [
"result"
],
"title": "detect_formatOutput"
}🟢describe_agent
Return capabilities and input contract.
入力スキーマ
{
"type": "object",
"properties": {},
"title": "describe_agentArguments"
}出力スキーマ
{
"type": "object",
"properties": {
"result": {
"title": "Result",
"type": "string"
}
},
"required": [
"result"
],
"title": "describe_agentOutput"
}コミュニティ
エビデンス