recap

name	recap
description	Recover the full shape of the available conversation as a structured map with reusable topic and axis labels, followed by an overall synthesis. Use only when the user invokes recap or asks for a recap or synthesis; never insert an unsolicited checkpoint.

🗺️ Think Recap

Use when: The discussion has lost its overall shape or needs a checkpoint. Default binding: The full available conversation, the same Binding as /on-conversation. Accepts: A compatible HACP Working Object or the declared default material. Effect: Reconstruct topics and axes with concise human labels, reuse supported labels, classify their contents and states, then synthesize relationships across them. Result: A navigable map whose topic and axis labels can be reused by binding cards, followed by a coherent account of where the thinking stands. Duration: One agent turn. Play it again at useful checkpoints. Limits: Preserve uncertainty and disagreement. Do not suggest another command, choose a direction, decide, plan, create technical identifiers, persist state, or create a file.

Flow

flowchart LR
    A["Selected conversation context"] --> B["Recover topics and axes"]
    B --> C["Name or reuse human labels"]
    C --> D["Classify content and states"]
    D --> E["Surface tensions and dependencies"]
    E --> F["Build navigable map"]
    F --> G["Write overall synthesis"]

Format

Begin the combo trace with > 🎯 **<binding>** → 🗺️ **RECAP**, followed by Map and Digest. Show the Conversation → Topics → Axes tree and the state of each axis. Give every topic and axis a concise label the user can repeat with on-topic or on-axis.

Add an output with → and presentation cards with +. Do not append suggested commands.