Length

Validates that the agent output's character or word count falls within specified bounds. Useful for enforcing minimum detail or capping verbose responses.

Config

Use Cases

• Minimum detail requirement — Ensure the agent provides substantive responses rather than one-word answers by setting a minimum word or character count.
• Conciseness enforcement — Cap response length to prevent overly verbose output that wastes tokens and frustrates users.
• Format compliance — Validate that summaries, titles, or short-form outputs stay within character limits (e.g. 280 characters for a tweet-style summary).
• Empty output detection — Catch cases where the agent returns an empty or near-empty response by requiring at least a minimum length.

Examples

Minimum word count

// Require at least 10 words
{
  "min": 10,
  "unit": "words"
}
// Output: "Yes" → fail (1 word)
// Output: "The product is available in three colors and ships within two days" → pass (12 words)

Character range

// Between 50 and 500 characters
{
  "min": 50,
  "max": 500,
  "unit": "chars"
}
// Output: "Short." → fail (6 chars)
// Output: (200 character response) → pass

Maximum word limit

// No more than 50 words
{
  "max": 50,
  "unit": "words"
}
// Output: (25 word summary) → pass
// Output: (80 word essay) → fail

Scoring

Returns 1.0 (pass) or 0.0 (fail). No intermediate scores. The reasoning message includes the actual count and the expected range.

Performance

Length performs a simple character or whitespace-split word count with no external API calls. Execution time is under 1ms regardless of output length, making it suitable for high-volume evaluation pipelines.