RegexForge
Deterministic regex synthesis from labeled examples. Zero LLM, proof matrix, backtracking audit.
我該用這個嗎
品質與安全性
A
根據工具定義與協定合規性的自動化分析。
上下文成本
~184Token(工具定義)
~1.3 KB典型回應大小
極小的注意力影響(128k 上下文的 0.14%)
這是每次將伺服器的工具載入模型上下文時所消耗的約略 token 數量。數量越高,可用於其他工作的注意力就越少。
安裝
一鍵安裝
將以下內容加入你的 `claude_desktop_config.json` 檔案:
{
"mcpServers": {
"regexforge": {
"url": "https://regexforge.jason-12c.workers.dev/mcp"
}
}
}遠端端點
https://regexforge.jason-12c.workers.dev/mcpstreamable-http它能做什麼
工具清單
工具(1)
🟢 唯讀🟡 寫入🔴 刪除⚪ 未知
⚪regexforge_synth(description, examples)
Synthesize a battle-tested regex from labeled examples. Input: optional natural-language description + 20 examples (10 positive, 10 negative, or any mix of ≥4). Output: regex + flags + test matrix proving every example is handled correctly + backtracking-risk analysis. Deterministic; no LLM at serve time.
輸入結構描述
{
"type": "object",
"properties": {
"description": {
"type": "string",
"description": "Optional natural-language description, used only for tie-breaking."
},
"examples": {
"type": "array",
"minItems": 2,
"maxItems": 100,
"items": {
"type": "object",
"required": [
"text",
"match"
],
"properties": {
"text": {
"type": "string",
"maxLength": 2048
},
"match": {
"type": "boolean",
"description": "true if the regex should match this string; false if it should NOT match."
}
}
}
}
},
"required": [
"examples"
]
}社群
證據
近期觀測
已驗證未記錄版本1 個工具
已驗證未記錄版本1 個工具