ishenwei/llm-wiki-agent
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat(tools): migrate to litellm for multi-provider support (OpenAI, Gemini, Claude)

Browse Source
This commit is contained in:
watsonk1998
2026-04-13 22:26:11 +08:00
parent d8ac6107bf
commit 818f8a2d15
5 changed files with 86 additions and 59 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
requirements.txt
View File

@@ -1,2 +1,2 @@
anthropic>=0.40.0
litellm>=1.0.0
networkx>=3.2

35
tools/build_graph.py
View File

@@ -25,7 +25,7 @@ import webbrowser
from pathlib import Path
from datetime import date
import anthropic
import os
try:
import networkx as nx
@@ -64,6 +64,23 @@ def read_file(path: Path) -> str:
return path.read_text(encoding="utf-8") if path.exists() else ""
def call_llm(prompt: str, model_env: str, default_model: str, max_tokens: int = 4096) -> str:
try:
from litellm import completion
except ImportError:
print("Error: litellm not installed. Run: pip install litellm")
import sys
sys.exit(1)
model = os.getenv(model_env, default_model)
response = completion(
model=model,
messages=[{"role": "user", "content": prompt}],
max_tokens=max_tokens
)
return response.choices[0].message.content
def sha256(text: str) -> str:
return hashlib.sha256(text.encode()).hexdigest()
@@ -143,8 +160,7 @@ def build_extracted_edges(pages: list[Path]) -> list[dict]:
def build_inferred_edges(pages: list[Path], existing_edges: list[dict], cache: dict) -> list[dict]:
"""Pass 2: Claude-inferred semantic relationships."""
client = anthropic.Anthropic()
"""Pass 2: API-inferred semantic relationships."""
new_edges = []
# Only process pages that changed since last run
@@ -172,12 +188,7 @@ def build_inferred_edges(pages: list[Path], existing_edges: list[dict], cache: d
content = read_file(p)[:2000] # truncate for context efficiency
src = page_id(p)
response = client.messages.create(
model="claude-haiku-4-5-20251001",
max_tokens=1024,
messages=[{
"role": "user",
"content": f"""Analyze this wiki page and identify implicit semantic relationships to other pages in the wiki.
prompt = f"""Analyze this wiki page and identify implicit semantic relationships to other pages in the wiki.
Source page: {src}
Content:
@@ -200,10 +211,8 @@ Rules:
- Do not repeat edges already in the extracted list
- Return empty array [] if no new relationships found
"""
}]
)
raw = response.content[0].text.strip()
raw = call_llm(prompt, "LLM_MODEL_FAST", "claude-3-5-haiku-latest", max_tokens=1024)
raw = raw.strip()
raw = re.sub(r"^```(?:json)?\s*", "", raw)
raw = re.sub(r"\s*```$", "", raw)

30
tools/ingest.py
View File

@@ -23,7 +23,7 @@ import re
from pathlib import Path
from datetime import date
import anthropic
import os
REPO_ROOT = Path(__file__).parent.parent
WIKI_DIR = REPO_ROOT / "wiki"
@@ -41,6 +41,22 @@ def read_file(path: Path) -> str:
return path.read_text(encoding="utf-8") if path.exists() else ""
def call_llm(prompt: str, max_tokens: int = 8192) -> str:
try:
from litellm import completion
except ImportError:
print("Error: litellm not installed. Run: pip install litellm")
sys.exit(1)
model = os.getenv("LLM_MODEL", "claude-3-5-sonnet-latest")
response = completion(
model=model,
messages=[{"role": "user", "content": prompt}],
max_tokens=max_tokens
)
return response.choices[0].message.content
def write_file(path: Path, content: str):
path.parent.mkdir(parents=True, exist_ok=True)
path.write_text(content, encoding="utf-8")
@@ -105,7 +121,7 @@ def ingest(source_path: str):
wiki_context = build_wiki_context()
schema = read_file(SCHEMA_FILE)
client = anthropic.Anthropic()
schema = read_file(SCHEMA_FILE)
prompt = f"""You are maintaining an LLM Wiki. Process this source document and integrate its knowledge into the wiki.
@@ -140,14 +156,8 @@ Return ONLY a valid JSON object with these fields (no markdown fences, no prose
}}
"""
print(" calling Claude API...")
response = client.messages.create(
model="claude-sonnet-4-6",
max_tokens=8192,
messages=[{"role": "user", "content": prompt}],
)
raw = response.content[0].text
print(f" calling API (model: ...)")
raw = call_llm(prompt, max_tokens=8192)
try:
data = parse_json_from_response(raw)
except (ValueError, json.JSONDecodeError) as e:

33
tools/lint.py
View File

@@ -21,7 +21,7 @@ from pathlib import Path
from collections import defaultdict
from datetime import date
import anthropic
import os
REPO_ROOT = Path(__file__).parent.parent
WIKI_DIR = REPO_ROOT / "wiki"
@@ -33,6 +33,22 @@ def read_file(path: Path) -> str:
return path.read_text(encoding="utf-8") if path.exists() else ""
def call_llm(prompt: str, model_env: str, default_model: str, max_tokens: int = 4096) -> str:
try:
from litellm import completion
except ImportError:
print("Error: litellm not installed. Run: pip install litellm")
sys.exit(1)
model = os.getenv(model_env, default_model)
response = completion(
model=model,
messages=[{"role": "user", "content": prompt}],
max_tokens=max_tokens
)
return response.choices[0].message.content
def all_wiki_pages() -> list[Path]:
return [p for p in WIKI_DIR.rglob("*.md")
if p.name not in ("index.md", "log.md", "lint-report.md")]
@@ -112,14 +128,8 @@ def run_lint():
rel = p.relative_to(REPO_ROOT)
pages_context += f"\n\n### {rel}\n{read_file(p)[:1500]}" # truncate long pages
client = anthropic.Anthropic()
print(" running semantic lint via Claude API...")
response = client.messages.create(
model="claude-sonnet-4-6",
max_tokens=3000,
messages=[{
"role": "user",
"content": f"""You are linting an LLM Wiki. Review the pages below and identify:
print(" running semantic lint via API...")
prompt = f"""You are linting an LLM Wiki. Review the pages below and identify:
1. Contradictions between pages (claims that conflict)
2. Stale content (summaries that newer sources have superseded)
3. Data gaps (important questions the wiki can't answer — suggest specific sources to find)
@@ -136,10 +146,7 @@ Return a markdown lint report with these sections:
Be specific — name the exact pages and claims involved.
"""
}]
)
semantic_report = response.content[0].text
semantic_report = call_llm(prompt, "LLM_MODEL", "claude-3-5-sonnet-latest", max_tokens=3000)
# Compose full report
report_lines = [

45
tools/query.py
View File

@@ -19,7 +19,7 @@ import argparse
from pathlib import Path
from datetime import date
import anthropic
import os
REPO_ROOT = Path(__file__).parent.parent
WIKI_DIR = REPO_ROOT / "wiki"
@@ -38,6 +38,22 @@ def write_file(path: Path, content: str):
print(f" saved: {path.relative_to(REPO_ROOT)}")
def call_llm(prompt: str, model_env: str, default_model: str, max_tokens: int = 4096) -> str:
try:
from litellm import completion
except ImportError:
print("Error: litellm not installed. Run: pip install litellm")
sys.exit(1)
model = os.getenv(model_env, default_model)
response = completion(
model=model,
messages=[{"role": "user", "content": prompt}],
max_tokens=max_tokens
)
return response.choices[0].message.content
def find_relevant_pages(question: str, index_content: str) -> list[Path]:
"""Extract linked pages from index that seem relevant to the question."""
# Pull all [[links]] and markdown links from index
@@ -64,7 +80,6 @@ def append_log(entry: str):
def query(question: str, save_path: str | None = None):
today = date.today().isoformat()
client = anthropic.Anthropic()
# Step 1: Read index
index_content = read_file(INDEX_FILE)
@@ -77,16 +92,10 @@ def query(question: str, save_path: str | None = None):
# If no keyword match, ask Claude to identify relevant pages from the index
if not relevant_pages or len(relevant_pages) <= 1:
print(" selecting relevant pages via Claude...")
selection_response = client.messages.create(
model="claude-haiku-4-5-20251001",
max_tokens=512,
messages=[{
"role": "user",
"content": f"Given this wiki index:\n\n{index_content}\n\nWhich pages are most relevant to answering: \"{question}\"\n\nReturn ONLY a JSON array of relative file paths (as listed in the index), e.g. [\"sources/foo.md\", \"concepts/Bar.md\"]. Maximum 10 pages."
}]
)
raw = selection_response.content[0].text.strip()
print(" selecting relevant pages via API...")
prompt = f"Given this wiki index:\n\n{index_content}\n\nWhich pages are most relevant to answering: \"{question}\"\n\nReturn ONLY a JSON array of relative file paths (as listed in the index), e.g. [\"sources/foo.md\", \"concepts/Bar.md\"]. Maximum 10 pages."
raw = call_llm(prompt, "LLM_MODEL_FAST", "claude-3-5-haiku-latest", max_tokens=512)
raw = raw.strip()
raw = re.sub(r"^```(?:json)?\s*", "", raw)
raw = re.sub(r"\s*```$", "", raw)
try:
@@ -108,12 +117,7 @@ def query(question: str, save_path: str | None = None):
# Step 4: Synthesize answer
print(f" synthesizing answer from {len(relevant_pages)} pages...")
response = client.messages.create(
model="claude-sonnet-4-6",
max_tokens=4096,
messages=[{
"role": "user",
"content": f"""You are querying an LLM Wiki to answer a question. Use the wiki pages below to synthesize a thorough answer. Cite sources using [[PageName]] wikilink syntax.
prompt = f"""You are querying an LLM Wiki to answer a question. Use the wiki pages below to synthesize a thorough answer. Cite sources using [[PageName]] wikilink syntax.
Schema:
{schema}
@@ -125,10 +129,7 @@ Question: {question}
Write a well-structured markdown answer with headers, bullets, and [[wikilink]] citations. At the end, add a ## Sources section listing the pages you drew from.
"""
}]
)
answer = response.content[0].text
answer = call_llm(prompt, "LLM_MODEL", "claude-3-5-sonnet-latest", max_tokens=4096)
print("\n" + "=" * 60)
print(answer)
print("=" * 60)