search for standard sites
pub-search.waow.tech
search
zig
blog
atproto
docs: fix READMEs with proper MCP setup commands
+29
-45
+10
README.md
+10
README.md
···
12
12
2. **backend** indexes content into SQLite FTS5 via [Turso](https://turso.tech), serves search API
13
13
3. **site** static frontend on Cloudflare Pages
14
14
15
+
## MCP server
16
+
17
+
search is also exposed as an MCP server for AI agents like Claude Code:
18
+
19
+
```bash
20
+
claude mcp add-json leaflet '{"type": "http", "url": "https://leaflet-search-by-zzstoatzz.fastmcp.app/mcp"}'
21
+
```
22
+
23
+
see [mcp/README.md](mcp/README.md) for local setup and usage details.
24
+
15
25
## api
16
26
17
27
```
+19
-45
mcp/README.md
+19
-45
mcp/README.md
···
2
2
3
3
MCP server for [Leaflet](https://leaflet.pub) - search decentralized publications on ATProto.
4
4
5
-
## installation
6
-
7
-
```bash
8
-
uv add leaflet-mcp
9
-
```
10
-
11
5
## usage
12
6
13
-
### as a CLI
7
+
### hosted (recommended)
14
8
15
9
```bash
16
-
leaflet-mcp
10
+
claude mcp add-json leaflet '{"type": "http", "url": "https://leaflet-search-by-zzstoatzz.fastmcp.app/mcp"}'
17
11
```
18
12
19
-
### with Claude Code
13
+
### local
20
14
21
-
Add to your `.claude/settings.json`:
15
+
run the MCP server locally with `uvx`:
22
16
23
-
```json
24
-
{
25
-
"mcpServers": {
26
-
"leaflet": {
27
-
"command": "uvx",
28
-
"args": ["leaflet-mcp"]
29
-
}
30
-
}
31
-
}
17
+
```bash
18
+
uvx --from git+https://github.com/zzstoatzz/leaflet-search#subdirectory=mcp leaflet-mcp
32
19
```
33
20
34
-
## tools
35
-
36
-
- `search(query, tag, limit)` - search documents and publications
37
-
- `get_document(uri)` - get full document content by AT-URI
38
-
- `find_similar(uri, limit)` - find semantically similar documents
39
-
- `get_tags()` - list available tags with counts
40
-
- `get_stats()` - get index statistics
41
-
- `get_popular(limit)` - get popular search queries
21
+
to add it to claude code as a local stdio server:
42
22
43
-
## example
23
+
```bash
24
+
claude mcp add leaflet -- uvx --from 'git+https://github.com/zzstoatzz/leaflet-search#subdirectory=mcp' leaflet-mcp
25
+
```
44
26
45
-
```python
46
-
from fastmcp.client import Client
47
-
from fastmcp.client.transports import SSETransport
27
+
## workflow
48
28
49
-
async with Client(transport=SSETransport("http://localhost:8000/sse")) as client:
50
-
# search for python articles
51
-
results = await client.call_tool("search", {"query": "python"})
29
+
1. **search** for documents by query or tag
30
+
2. **get_document** to retrieve full content by AT-URI
52
31
53
-
# get full content of first result
54
-
if results.data:
55
-
doc = await client.call_tool("get_document", {"uri": results.data[0].uri})
56
-
print(doc.data.content)
32
+
```
33
+
search("space station") → [{uri: "at://...", title: "...", snippet: "..."}]
34
+
get_document("at://...") → {title: "...", content: "full article text..."}
57
35
```
58
36
59
37
## development
60
38
61
39
```bash
62
-
# install dev dependencies
63
-
uv sync --group dev
64
-
65
-
# run tests
40
+
git clone https://github.com/zzstoatzz/leaflet-search
41
+
cd leaflet-search/mcp
42
+
uv sync
66
43
uv run pytest
67
-
68
-
# format
69
-
uv run ruff format
70
44
```