Search Resources

kdeps provides two native search executors compiled into the binary: searchLocal for local file search and searchWeb for web search.

Where it runs

Both workflow mode and agent mode. In workflow mode each executor runs as a DAG step. In agent mode, the workflow containing these resources runs as a single callable tool.

---

searchLocal

The searchLocal executor walks a local directory and returns matching files by filename glob pattern and/or content keyword.

Configuration

# resources/search.yaml
searchLocal:
  path: "/data/documents"    # required: directory to search
  query: "invoice total"     # optional: keyword in file contents
  glob: "*.txt"              # optional: filename pattern
  limit: 10                  # optional: max results (0 = unlimited)

Field	Type	Required	Default	Description
path	string	yes	-	Directory to search recursively
query	string	no	-	Case-insensitive keyword to find in file contents
glob	string	no	-	Filename glob pattern (e.g. *.md, report_*.csv)
limit	integer	no	0	Max results (0 = unlimited)

When both query and glob are set, a file must match both to be included.

Output

Key	Type	Description
results	array	List of matching file objects
count	integer	Number of results
path	string	The search root used
json	string	Full result as JSON string

Each result object:

Key	Type	Description
path	string	Full file path
name	string	Filename
size	integer	File size in bytes
isDir	bool	Always false

Examples

Find all Markdown files:

# resources/find-docs.yaml
actionId: findDocs
searchLocal:
  path: "/workspace/docs"
  glob: "*.md"

Find files containing a keyword:

# resources/find-invoices.yaml
actionId: findInvoices
searchLocal:
  path: "/data/uploads"
  query: "overdue"
  limit: 20

Feed results into an LLM:

# resources/find-files.yaml
actionId: findFiles
searchLocal:
  path: "/data/reports"
  query: "{{ get('query') }}"

---
actionId: answer
requires: [findFiles]
chat:
  model: llama3.2:1b
  prompt: "Files found: {{ output('findFiles').results }}. Summarize."

---

searchWeb

The searchWeb executor queries the web and returns structured results. The default provider is DuckDuckGo -- no connection or API key required. Paid providers (Brave, Bing, Tavily) require a named connection in workflow.yaml settings.

Global Named Connections (paid providers)

API keys belong in ~/.kdeps/config.yaml, not inline in resource files or workflow.yaml:

# ~/.kdeps/config.yaml
search_connections:
  brave:
    apiKey: "${BRAVE_API_KEY}"
  bing:
    apiKey: "${BING_API_KEY}"
  tavily:
    apiKey: "${TAVILY_API_KEY}"

Configuration

# resources/example.yaml
searchWeb:
  query: "{{ get('query') }}"     # required
  provider: ddg                    # optional: ddg (default) | brave | bing | tavily
  connectionName: brave            # required for brave/bing/tavily; references settings.searchConnections
  maxResults: 5                    # optional, default 5
  timeout: 15                      # optional, seconds, default 15

Field	Type	Required	Default	Description
query	string	yes	-	Search query
provider	string	no	ddg	Search provider: ddg, brave, bing, tavily
connectionName	string	no*	-	Named connection from settings.searchConnections
maxResults	integer	no	5	Maximum number of results
timeout	integer	no	15	HTTP request timeout in seconds

*Required when provider is brave, bing, or tavily.

Providers

Provider	Value	Connection required
DuckDuckGo	ddg	No
Brave Search	brave	Yes
Bing	bing	Yes
Tavily	tavily	Yes

Output

output('search').results    # list of result objects
output('search').count      # number of results returned
output('search').query      # the query string
output('search').provider   # provider used
output('search').json       # JSON string of the full result

Result object fields:

Field	Type	Description
title	string	Page title
url	string	Page URL
snippet	string	Description or summary snippet

Examples

DuckDuckGo (no connection needed):

# resources/search.yaml
actionId: search
searchWeb:
  query: "{{ get('query') }}"
  maxResults: 5

Brave Search:

# ~/.kdeps/config.yaml
search_connections:
  brave:
    apiKey: "${BRAVE_API_KEY}"

# resources/search.yaml
actionId: search
searchWeb:
  query: "{{ get('query') }}"
  provider: brave
  connectionName: brave
  maxResults: 10

Feed results into an LLM:

# resources/search.yaml
actionId: search
searchWeb:
  query: "{{ get('query') }}"

---
actionId: answer
requires: [search]
chat:
  model: llama3.2:1b
  prompt: |
    Answer based on these results:
    {% for r in output('search').results %}
    - {{ r.title }}: {{ r.snippet }}
    {% endfor %}
    Question: {{ get('query') }}

Error handling

Error	Cause
query is required	Empty query string
connectionName required for brave provider	No connection set for Brave
connectionName required for bing provider	No connection set for Bing
connectionName required for tavily provider	No connection set for Tavily
unknown provider "x"	Invalid provider value

Environment variable overrides

Variable	Description
KDEPS_DDG_URL	Override DuckDuckGo base URL
KDEPS_BRAVE_URL	Override Brave API base URL
KDEPS_BING_URL	Override Bing API base URL
KDEPS_TAVILY_URL	Override Tavily API base URL

---

See Also

• Embedding Resource - SQLite keyword store for on-prem RAG
• Scraper Resource - Fetch URL content to feed into search pipelines
• LLM Resource - Use search results as context for chat resources