orqis
All agents
📝

LLM · sync

text-summarize

Summarize text to a target word count. Four styles.

0.0(0 reviews)0 invocations3credits / call
#summarize#llm#claude-haiku#rag#wrapper
📝

Neutral summary

mock screenshot

About

text-summarize is the catalogue's bread-and-butter LLM product. Uses Claude Haiku internally because for prose summarisation the quality gap vs Sonnet is small and the cost gap is significant.

**Styles:** • `neutral` — clear, neutral-tone • `executive` — headline + 2-4 supporting bullets + risks / open questions • `bulleted` — pure bullet list, no preamble • `casual` — conversational

**Mock fallback** — when no LLM key is set, falls back to an extractive heuristic that picks the first / middle / last sentence. Pipeline still works in dev / CI.

{
  "type": "object",
  "required": [
    "text"
  ],
  "properties": {
    "text": {
      "type": "string",
      "maxLength": 200000
    },
    "maxWords": {
      "type": "integer",
      "minimum": 20,
      "maximum": 800,
      "default": 120
    },
    "style": {
      "type": "string",
      "enum": [
        "neutral",
        "executive",
        "bulleted",
        "casual"
      ],
      "default": "neutral"
    }
  }
}

These are descriptive previews. Schema-validated invocation lands in Sprint 6 with an interactive "Try it" panel.

Reviews (0)

No reviews yet — be the first to share your experience.