Built-in tools for string manipulation operations.
These tools provide common string functionality that AI agents often need:
- Text transformation (uppercase, lowercase, capitalize)
- String analysis (length, count occurrences)
- String operations (replace, split, join, trim)
- Pattern matching and extraction
- String validation
Usage
agent = Nous.new("lmstudio:qwen3-vl-4b-thinking-mlx",
tools: [
&StringTools.string_length/2,
&StringTools.replace_text/2,
&StringTools.split_text/2
]
)
{:ok, result} = Nous.run(agent, "How many characters in 'Hello World'?")Summary
Types
Tool arguments exactly as the model produced them: JSON object keys stay strings, values are unvalidated. Every function below reads what it needs and falls back to a default for anything missing or of the wrong type.
Run context. These tools are pure and ignore it, but they are registered
with takes_ctx: true, so it is always the first argument.
Tool result: an atom-keyed map that echoes the inputs it used alongside the computed fields, so the model can see what it actually asked for.
Functions
Capitalize the first letter of each word.
Check if a string contains a substring.
Count occurrences of a substring in a string.
Check if a string ends with a suffix.
Extract numbers from a string.
Extract words from a string.
Check if a string is a palindrome.
Join a list of strings with a delimiter.
Pad a string to a specific length.
Repeat a string N times.
Replace all occurrences of a pattern in a string.
Reverse a string.
Split a string into parts based on a delimiter.
Check if a string starts with a prefix.
Get the length of a string.
Extract a substring from a string.
Convert text to lowercase.
Convert text to uppercase.
Trim whitespace from a string.
Types
Tool arguments exactly as the model produced them: JSON object keys stay strings, values are unvalidated. Every function below reads what it needs and falls back to a default for anything missing or of the wrong type.
@type ctx() :: Nous.RunContext.t() | nil
Run context. These tools are pure and ignore it, but they are registered
with takes_ctx: true, so it is always the first argument.
Tool result: an atom-keyed map that echoes the inputs it used alongside the computed fields, so the model can see what it actually asked for.
Functions
Capitalize the first letter of each word.
Arguments
- text: The text to capitalize
- mode: "first" (first letter only), "words" (each word), "sentences" (each sentence)
Check if a string contains a substring.
Arguments
- text: The text to search in
- pattern: The pattern to search for
- case_sensitive: Whether to match case (default: true)
Count occurrences of a substring in a string.
Arguments
- text: The text to search in
- pattern: The pattern to count
- case_sensitive: Whether to match case (default: true)
Check if a string ends with a suffix.
Arguments
- text: The text to check
- suffix: The suffix to check for
- case_sensitive: Whether to match case (default: true)
Extract numbers from a string.
Arguments
- text: The text to extract numbers from
Extract words from a string.
Arguments
- text: The text to extract words from
- min_length: Minimum word length (default: 1)
Check if a string is a palindrome.
Arguments
- text: The text to check
- ignore_case: Whether to ignore case (default: true)
- ignore_spaces: Whether to ignore spaces (default: true)
Join a list of strings with a delimiter.
Arguments
- parts: List of strings to join (comma-separated string)
- delimiter: The delimiter to use (default: " ")
Pad a string to a specific length.
Arguments
- text: The text to pad
- length: Target length
- padding: Character to pad with (default: " ")
- side: "left", "right", or "both" (default: "right")
Repeat a string N times.
Arguments
- text: The text to repeat
- times: Number of times to repeat (max 100)
Replace all occurrences of a pattern in a string.
Arguments
- text: The original text
- pattern: The text to find
- replacement: The text to replace with
- case_sensitive: Whether to match case (default: true)
Reverse a string.
Arguments
- text: The text to reverse
Split a string into parts based on a delimiter.
Arguments
- text: The text to split
- delimiter: The delimiter to split on (default: " ")
- trim: Whether to trim whitespace from parts (default: false)
- remove_empty: Whether to remove empty strings (default: false)
Check if a string starts with a prefix.
Arguments
- text: The text to check
- prefix: The prefix to check for
- case_sensitive: Whether to match case (default: true)
Get the length of a string.
Arguments
- text: The string to measure
Extract a substring from a string.
Arguments
- text: The original text
- start: Starting position (0-indexed)
- length: Number of characters to extract (optional, extracts to end if not provided)
Convert text to lowercase.
Arguments
- text: The text to convert
Convert text to uppercase.
Arguments
- text: The text to convert
Trim whitespace from a string.
Arguments
- text: The text to trim
- side: "both" (default), "left", "right"