Patent Disclosure Skill Trigger Keywords: Complete List and Implementation

The Patent Disclosure skill activates when a voice assistant detects specific trigger keywords including "patent", "disclosure", "patent search", and eight other terms defined in the skill's manifest.json configuration.

The handsomestWei/patent-disclosure-skill repository implements a voice-activated patent analysis system that responds to predefined utterance patterns. Understanding the trigger keywords for the patent disclosure skill is essential for developers integrating this capability into voice assistants or extending its natural language processing capabilities.

Where Trigger Keywords Are Configured

The authoritative list of trigger keywords resides in the skill's manifest file. In skills/patent-disclosure/manifest.json, the "triggers" field defines the exact terms that route user requests to this skill.

When the voice assistant's NLP engine parses incoming speech, it references this manifest to determine skill routing. Any match between the normalized user input and the trigger list causes the request to be forwarded to the skill's entry point.

Complete List of Patent Disclosure Skill Trigger Keywords

The manifest defines eleven distinct trigger keywords and phrases that activate the skill:

• patent
• disclosure
• patent disclosure
• patent info
• patent search
• patent analysis
• patent review
• patent summary
• patent report
• patent filing
• patent details

These terms function individually or in combination with other words. For example, utterances like "show me patent details" or "I need a patent search" both trigger the skill because they contain the core keywords "patent details" and "patent search" respectively.

How Trigger Detection Works in the Source Code

Once a keyword match occurs in the manifest layer, the request routes to skills/patent-disclosure/tools/cad_scan.py, which serves as the skill's primary entry point. The following Python logic demonstrates how the trigger detection mechanism normalizes input and checks against the keyword set:

def is_patent_disclosure_trigger(text: str) -> bool:
    triggers = {
        "patent", "disclosure", "patent disclosure", "patent info",
        "patent search", "patent analysis", "patent review",
        "patent summary", "patent report", "patent filing",
        "patent details"
    }
    # Normalise to lower-case and look for any trigger word

    lowered = text.lower()
    return any(trigger in lowered for trigger in triggers)

# Sample usage

user_utterance = "Can you give me a patent summary for my invention?"
if is_patent_disclosure_trigger(user_utterance):
    print("Routing to Patent Disclosure skill…")

The function converts input to lowercase to ensure case-insensitive matching, then checks if any trigger keyword exists as a substring within the user utterance. This approach allows for flexible natural language variations while maintaining precise keyword boundaries.

Testing Trigger Keyword Routing

The repository includes unit tests in skills/patent-disclosure/tests/test_cad_step_scan.py that verify the trigger mechanism correctly identifies valid keywords and rejects unrelated utterances. These tests ensure that modifications to the manifest or detection logic do not break the routing functionality.

Developers should run these tests when adding new trigger keywords or modifying the normalization logic in cad_scan.py to prevent regression in skill activation accuracy.

Summary

• The Patent Disclosure skill activates upon detecting eleven specific trigger keywords defined in manifest.json.
• Key trigger terms include "patent", "disclosure", "patent search", and "patent analysis".
• The entry point at skills/patent-disclosure/tools/cad_scan.py handles requests after manifest-level routing occurs.
• Input normalization uses lowercase conversion and substring matching to support natural language variations.
• Unit tests in test_cad_step_scan.py validate trigger detection accuracy.

Frequently Asked Questions

How do I add new trigger keywords to the Patent Disclosure skill?

Add new terms to the "triggers" array in skills/patent-disclosure/manifest.json, then update the triggers set in cad_scan.py to include the new keywords. Run the tests in test_cad_step_scan.py to verify the skill correctly routes requests containing the new terms.

Does the Patent Disclosure skill require exact keyword matching?

No, the skill uses substring matching after lowercasing the input. This means "patent summary" will trigger the skill even when embedded in longer phrases like "generate a quick patent summary for me."

What happens when multiple trigger keywords appear in one utterance?

The skill activates as long as at least one trigger keyword matches. Multiple keywords do not change the routing behavior; the request still processes through cad_scan.py following the standard patent analysis workflow.

Where is the trigger keyword list officially defined?

The authoritative list resides in skills/patent-disclosure/manifest.json under the triggers field, which the voice assistant's skill router references to determine when to invoke this patent analysis capability.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}