Semble: поиск по коду для AI-агентов за миллисекунды

claude mcp add semble -s user -- uvx --from "semble[mcp]" semble

uv tool install semble   # Установка через uv (рекомендуется)
pip install semble       # Или через pip

## Code Search

Use `semble search` to find code by describing what it does or naming a symbol/identifier, instead of grep:

```bash
semble search "authentication flow" ./my-project
semble search "save_pretrained" ./my-project
semble search "save model to disk" ./my-project --top-k 10
```

The index is built on first run (and cached for subsequent runs) and invalidated automatically when files change.

Use `--content docs` to search documentation and prose, `--content config` for config files (yaml, toml, etc.), or `--content all` to search code, docs, and config:

```bash
semble search "deployment guide" ./my-project --content docs
semble search "database host port" ./my-project --content config
semble search "authentication" ./my-project --content all
```

Use `semble find-related` to discover code similar to a known location (pass `file_path` and `line` from a prior search result):

```bash
semble find-related src/auth.py 42 ./my-project
```

`path` defaults to the current directory when omitted; git URLs are accepted.

If `semble` is not on `$PATH`, use `uvx --from "semble[mcp]" semble` in its place.

### Workflow

1. Start with `semble search` to find relevant chunks. The index is built and cached automatically.
2. Use `--content docs` for documentation, `--content config` for config files, or `--content all` for everything.
3. Inspect full files only when the returned chunk does not give enough context.
4. Optionally use `semble find-related` with a promising result's `file_path` and `line` to discover related implementations.
5. Use grep only when you need exhaustive literal matches or quick confirmation of an exact string.

semble init   # Claude Code → .claude/agents/semble-search.md

uv tool upgrade semble         # через uv
uv cache clean semble          # для пользователей MCP (после этого перезапустите MCP-клиент)
pip install --upgrade semble   # через pip

claude mcp add semble -s user -- uvx --from "semble[mcp]" semble

{
  "mcpServers": {
    "semble": {
      "command": "uvx",
      "args": ["--from", "semble[mcp]", "semble"]
    }
  }
}

[mcp_servers.semble]
command = "uvx"
args = ["--from", "semble[mcp]", "semble"]

{
  "mcp": {
    "semble": {
      "type": "local",
      "command": ["uvx", "--from", "semble[mcp]", "semble"]
    }
  }
}

{
  "servers": {
    "semble": {
      "command": "uvx",
      "args": ["--from", "semble[mcp]", "semble"]
    }
  }
}

{
  "mcpServers": {
    "semble": {
      "command": "uvx",
      "args": ["--from", "semble[mcp]", "semble"]
    }
  }
}

{
  "mcpServers": {
    "semble": {
      "command": "uvx",
      "args": ["--from", "semble[mcp]", "semble"]
    }
  }
}

{
  "mcpServers": {
    "semble": {
      "command": "uvx",
      "args": ["--from", "semble[mcp]", "semble"]
    }
  }
}

{
  "mcpServers": {
    "semble": {
      "command": "uvx",
      "args": ["--from", "semble[mcp]", "semble"]
    }
  }
}

{
  "context_servers": {
    "semble": {
      "command": "uvx",
      "args": ["--from", "semble[mcp]", "semble"]
    }
  }
}

# .sembleignore
generated/     # исключить директорию generated
*.pb.go        # исключить Go protobuf-файлы

# .sembleignore
!*.proto       # включить Protobuf-файлы
!*.cob         # включить COBOL-файлы

semble savings           # сводка по периодам
semble savings --verbose # также показать разбивку по типам вызовов

  Semble Token Savings
  ════════════════════════════════════════════════════════════════
  Period        Calls   Savings
  ────────────────────────────────────────────────────────────────
  Today         42      [███████████████░]  ~58.4k tokens (95%)
  Last 7 days   287     [██████████████░░]  ~312.4k tokens (90%)
  All time      1.4k    [██████████████░░]  ~1.2M tokens (89%)

from semble import ContentType, SembleIndex

# Индексировать локальную директорию (только код, по умолчанию)
index = SembleIndex.from_path("./my-project")

# Индексировать документацию и текст (markdown, rst и т.д.)
index = SembleIndex.from_path("./my-project", content=ContentType.DOCS)

# Индексировать всё (код, документацию и конфиги)
index = SembleIndex.from_path("./my-project", content=[ContentType.CODE, ContentType.DOCS, ContentType.CONFIG])

# Индексировать код и документацию вместе
index = SembleIndex.from_path("./my-project", content=[ContentType.CODE, ContentType.DOCS])

# Индексировать удалённый git-репозиторий
index = SembleIndex.from_git("https://github.com/MinishLab/model2vec")

# Поиск по индексу на естественном языке или по фрагменту кода
results = index.search("save model to disk", top_k=3)

# Найти код, похожий на конкретный результат
related = index.find_related(results[0], top_k=3)

# Каждый результат предоставляет доступ к найденному фрагменту
result = results[0]
result.chunk.file_path   # "model2vec/model.py"
result.chunk.start_line  # 127
result.chunk.end_line    # 150
result.chunk.content     # "def save_pretrained(self, path: PathLike, ..."