ResonanceEngine

What This Skill Does

ResonanceEngine is a sophisticated conversational analysis tool designed for OpenClaw agents. It functions as a lightweight, algorithmic layer that operates directly on the local thread, requiring zero external API calls. By measuring over 15 distinct micro-signals—ranging from sentiment velocity and hedging patterns to word-choice mirroring—ResonanceEngine calculates four critical frequencies: Engagement, Trust, Decision, and Style Match. Instead of relying on heavy machine learning inference, it uses heuristic text analysis to map the user's conversational state, allowing your agent to dynamically adjust its tone, depth, and persuasiveness in real-time. By matching the user's natural conversational frequency, your bot creates a sense of deep rapport that dramatically improves retention, conversion rates, and the overall quality of user interaction.

Installation

To integrate ResonanceEngine into your OpenClaw environment, use the internal skill manager. Run the following command in your terminal or via the OpenClaw management console:

clawhub install openclaw/skills/skills/jcools1977/openpaw

Once installed, ensure you have the openpaw library accessible in your agent's runtime directory to begin importing the ResonanceEngine class.

Use Cases

1. Sales Conversion: Detects when a user is in a 'Decision' state and triggers an urgent, value-driven call to action. It identifies hesitation markers and automatically pivots to objection-handling mode.
2. Support & Resolution: Analyzes frustration signals (sentiment velocity drops) and triggers empathy-first response patterns to prevent escalation.
3. Long-term Companion Bots: Maintains 'Style Match' consistency over hundreds of turns, ensuring the bot's communication style evolves in lockstep with the user's personality.

Example Prompts

1. "Analyze the current conversation thread for 'Decision' frequency and tell me if I should push for a sale or ask a discovery question."
2. "ResonanceEngine, give me a tone profile for the user's last three messages and suggest a response that maximizes trust."
3. "If my resonance level is currently 'WEAK', suggest a pivot strategy to get the user re-engaged without sounding pushy."

Tips & Limitations

Tips:

• Always pipe the result.recommendation.to_prompt_injection() output directly into your system prompt to ensure the agent follows the advice precisely.
• Use the resonance_level output to adjust the verbosity of your bot; for instance, shorten responses during 'WEAK' resonance to avoid overwhelming the user.

Limitations:

• The engine is limited to the local text stream provided to it; it cannot see external user telemetry or account history outside the conversation object.
• It is optimized for English; performance may degrade with highly fragmented or multi-lingual input.