Files
DocsGPT/application/agents/tools/brave.py
T
Alex 636d4d35d1 feat(prompts): revamp preset prompts, tool naming, and prompt templating
Rewrite the default/creative/strict presets (classic + agentic) into
structured sections: grounding and cite-by-title guidance, insufficient-
context behavior, current date, respond-in-user-language, scoped mermaid
usage, an untrusted-content guardrail, and a conditional XML-tagged
document context block. A memory directory listing is injected at render
time via the template prefetch mechanism so the model starts oriented
without burning a tool call.

Fixes along the way:
- Agentic preset swap was dead code: _get_prompt_content cached the
  classic preset before create_agent's swap check ran, so agentic and
  research agents always got the classic preset. The swap now happens
  inside _get_prompt_content.
- Jinja autoescape corrupted document content in custom prompts
  (< -> &lt;); prompts are not HTML, autoescape is now off.
- Literal {summaries} leaked into the prompt when no docs were
  retrieved; the placeholder is now stripped.
- Agentic/research prompts referenced tool names from a dropped naming
  scheme (search_internal, reason_think); they now reference the real
  names (search, reason).
- The strict preset told the model to "be very creative and use your
  imagination" right after "never make up information".
- extract_tool_usages recorded intermediate attribute chains as
  bare-tool usages, which meant "run all actions" at prefetch; only
  maximal chains are recorded now.
- Headless runs retrieved docs but never rendered them into the
  prompt; the prompt is now rendered like the streaming path.
- Default tools were unreachable by name in prompt templates
  (prefetch results were keyed by synthetic id only); defaults now
  claim the name key unless an explicit row shadows it.

Tool layer: memory/notes/todo actions are namespaced (memory_view,
note_overwrite, todo_create, ...) with legacy unprefixed names still
accepted via prefix stripping; duplicate action names across tools are
disambiguated with the owning tool's name instead of numeric suffixes;
thin tool descriptions rewritten (brave, duckduckgo, telegram, ntfy,
cryptoprice, read_webpage, internal_search, think).

Docs are now wrapped per chunk in <document index>/<source>/<content>
tags for citation-by-title support.
2026-06-11 11:18:46 +01:00

199 lines
6.2 KiB
Python

import logging
import requests
from application.agents.tools.base import Tool
logger = logging.getLogger(__name__)
class BraveSearchTool(Tool):
"""
Brave Search
A tool for performing web and image searches using the Brave Search API.
Requires an API key for authentication.
"""
def __init__(self, config):
self.config = config
self.token = config.get("token", "")
self.base_url = "https://api.search.brave.com/res/v1"
def execute_action(self, action_name, **kwargs):
actions = {
"brave_web_search": self._web_search,
"brave_image_search": self._image_search,
}
if action_name in actions:
return actions[action_name](**kwargs)
else:
raise ValueError(f"Unknown action: {action_name}")
def _web_search(
self,
query,
country="ALL",
search_lang="en",
count=10,
offset=0,
safesearch="off",
freshness=None,
result_filter=None,
extra_snippets=False,
summary=False,
):
"""
Performs a web search using the Brave Search API.
"""
logger.debug("Performing Brave web search for: %s", query)
url = f"{self.base_url}/web/search"
params = {
"q": query,
"country": country,
"search_lang": search_lang,
"count": min(count, 20),
"offset": min(offset, 9),
"safesearch": safesearch,
}
if freshness:
params["freshness"] = freshness
if result_filter:
params["result_filter"] = result_filter
if extra_snippets:
params["extra_snippets"] = 1
if summary:
params["summary"] = 1
headers = {
"Accept": "application/json",
"Accept-Encoding": "gzip",
"X-Subscription-Token": self.token,
}
response = requests.get(url, params=params, headers=headers, timeout=100)
if response.status_code == 200:
return {
"status_code": response.status_code,
"results": response.json(),
"message": "Search completed successfully.",
}
else:
return {
"status_code": response.status_code,
"message": f"Search failed with status code: {response.status_code}.",
}
def _image_search(
self,
query,
country="ALL",
search_lang="en",
count=5,
safesearch="off",
spellcheck=False,
):
"""
Performs an image search using the Brave Search API.
"""
logger.debug("Performing Brave image search for: %s", query)
url = f"{self.base_url}/images/search"
params = {
"q": query,
"country": country,
"search_lang": search_lang,
"count": min(count, 100), # API max is 100
"safesearch": safesearch,
"spellcheck": 1 if spellcheck else 0,
}
headers = {
"Accept": "application/json",
"Accept-Encoding": "gzip",
"X-Subscription-Token": self.token,
}
response = requests.get(url, params=params, headers=headers, timeout=100)
if response.status_code == 200:
return {
"status_code": response.status_code,
"results": response.json(),
"message": "Image search completed successfully.",
}
else:
return {
"status_code": response.status_code,
"message": f"Image search failed with status code: {response.status_code}.",
}
def get_actions_metadata(self):
return [
{
"name": "brave_web_search",
"description": (
"Search the web with Brave Search. Returns result titles, "
"URLs, and snippets. Use it for current events or "
"information not found in the user's documents."
),
"parameters": {
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "The search query (max 400 characters, 50 words)",
},
"search_lang": {
"type": "string",
"description": "The search language preference (default: en)",
},
"freshness": {
"type": "string",
"description": "Time filter for results (pd: last 24h, pw: last week, pm: last month, py: last year)",
},
},
"required": ["query"],
"additionalProperties": False,
},
},
{
"name": "brave_image_search",
"description": (
"Search for images with Brave Search. Returns image "
"titles, page URLs, and thumbnail URLs."
),
"parameters": {
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "The search query (max 400 characters, 50 words)",
},
"count": {
"type": "integer",
"description": "Number of results to return (max 100, default: 5)",
},
},
"required": ["query"],
"additionalProperties": False,
},
},
]
def get_config_requirements(self):
return {
"token": {
"type": "string",
"label": "API Key",
"description": "Brave Search API key for authentication",
"required": True,
"secret": True,
"order": 1,
},
}