Add Stack Exchange ZIMs and wire Kiwix search into MCP server

- kiwix_download.sh: add Ask Ubuntu, Super User, Unix & Linux SE, Server Fault
- mcp_server.py: add kiwix_search tool — models can now query all offline ZIMs
- local-ai-setup.sh: pass KIWIX_URL env to MCP container, add kiwix dependency

https://claude.ai/code/session_012gDnantBmFTWZGCiKyjazx
This commit is contained in:
Claude
2026-03-21 19:58:57 +00:00
parent f0eca342a7
commit 15eccd7e46
3 changed files with 47 additions and 4 deletions
+22
View File
@@ -16,6 +16,7 @@ GITEA_URL = os.getenv("GITEA_URL", "http://gitea:3000")
GITEA_TOKEN = os.getenv("GITEA_TOKEN", "")
GITHUB_TOKEN= os.getenv("GITHUB_TOKEN","")
RAG_URL = os.getenv("RAG_URL", "http://rag-server:8001")
KIWIX_URL = os.getenv("KIWIX_URL", "http://kiwix:80")
mcp = FastMCP("local-dev-tools")
@@ -180,6 +181,27 @@ def github_api(method: str, endpoint: str, body: str = "") -> str:
except Exception:
return r.text
# ── Kiwix search ──────────────────────────────────────────────────────────────
import re as _re
@mcp.tool()
def kiwix_search(query: str, books: str = "") -> str:
"""Search offline ZIM content via Kiwix (Wikipedia, Stack Overflow, etc).
books: optional comma-separated book names to filter, e.g. 'stackoverflow.com_en_all'.
Leave empty to search all ZIMs."""
params = {"pattern": query, "lang": ""}
if books:
params["books"] = books
try:
r = httpx.get(f"{KIWIX_URL}/search", params=params, timeout=15)
# strip HTML tags to get plain text
text = _re.sub(r"<[^>]+>", " ", r.text)
text = _re.sub(r"&[a-zA-Z]+;", " ", text)
text = _re.sub(r"\s{2,}", "\n", text).strip()
return text[:6000] or "(no results)"
except Exception as e:
return f"[kiwix error] {e}"
# ── RAG ingest ────────────────────────────────────────────────────────────────
@mcp.tool()
def ingest_repo(url: str, name: str = "", branch: str = "main") -> str: