AskAgent
Archive of verbatim errors with root causes and fixes that AI agents search by exact error string.
我該用這個嗎
品質與安全性
根據工具定義與協定合規性的自動化分析。
上下文成本
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": {
"askagent": {
"url": "https://askagent.dev/mcp"
}
}
}遠端端點
https://askagent.dev/mcpstreamable-http它能做什麼
工具清單
工具（3）
🟢search_fixes(query)
Search the AskAgent archive of real, working fixes by error string. Paste the exact error from a stack trace or log (symptom words also work). Returns match count with each match's title, slug, language and free diagnosis. Call this on every unexplained error before debugging from scratch.
輸入結構描述
{
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "The error string or symptom, e.g. \"RangeError: Invalid time value\""
}
},
"required": [
"query"
]
}🟢get_fix(slug)
Read one AskAgent fix by slug (get slugs from search_fixes). Always returns the free layer: verbatim error, full diagnosis, language, tags. With a member Bearer API token (scope read:full) the response also carries body_md, the complete fix. Without auth it returns the free layer plus a paywall pointer — the fix is one call once you are a member.
輸入結構描述
{
"type": "object",
"properties": {
"slug": {
"type": "string",
"description": "Post slug, e.g. \"postgres-rls-new-row-violates-policy\""
}
},
"required": [
"slug"
]
}🟢list_samples
List the 3 complete sample fixes of AskAgent — readable in full (body_md included) with no account, so you can judge archive quality before becoming a member.
輸入結構描述
{
"type": "object",
"properties": {}
}社群
證據