Skip to content

Add Ghost Protocol: Stealth & Perception plugin#191

Open
AATheBuilder wants to merge 1 commit intoagent0ai:mainfrom
AATheBuilder:add-ghost-protocol
Open

Add Ghost Protocol: Stealth & Perception plugin#191
AATheBuilder wants to merge 1 commit intoagent0ai:mainfrom
AATheBuilder:add-ghost-protocol

Conversation

@AATheBuilder
Copy link
Copy Markdown
Contributor

Professional OpSec guard with active stealth optimization and security tool output parsing.

@a0-bot
Copy link
Copy Markdown

a0-bot bot commented Apr 2, 2026

Plugin submission validation failed

ERROR: PR must modify exactly one plugin folder under plugins/

@3clyp50
Copy link
Copy Markdown
Collaborator

3clyp50 commented Apr 7, 2026

Hello @AATheBuilder I went through this plugin's code as well, and I just have a request for this one. The plugin is fine, but do we really need to inject the "silent" prompt in the repairable exception? Just a heads up, if maybe it's possible to have it in markdown in a prompts folder, to make review easier for everyone.

File: extensions/python/tool_execute_before/wraith_optimizer.py → lines 45–52

        agent.set_data("_wraith_eval_active", True)
        
        raise RepairableException(
            f"\n[WRAITH EVALUATION REQUIRED (Loudness {loudness_score}/10)]\n"
            f"The command '{command}' is too loud for professional stealth operations.\n"
            "MANDATORY ACTION: Propose 2-3 SILENT alternatives (e.g. passive OSINT, "
            "native OS binaries, or service-specific log analysis) before proceeding.\n"
            "To bypass, add '--force-loud' to your command."

@AATheBuilder
Copy link
Copy Markdown
Contributor Author

Hello @AATheBuilder I went through this plugin's code as well, and I just have a request for this one. The plugin is fine, but do we really need to inject the "silent" prompt in the repairable exception? Just a heads up, if maybe it's possible to have it in markdown in a prompts folder, to make review easier for everyone.

File: extensions/python/tool_execute_before/wraith_optimizer.py → lines 45–52

        agent.set_data("_wraith_eval_active", True)
        
        raise RepairableException(
            f"\n[WRAITH EVALUATION REQUIRED (Loudness {loudness_score}/10)]\n"
            f"The command '{command}' is too loud for professional stealth operations.\n"
            "MANDATORY ACTION: Propose 2-3 SILENT alternatives (e.g. passive OSINT, "
            "native OS binaries, or service-specific log analysis) before proceeding.\n"
            "To bypass, add '--force-loud' to your command."

On it!

@AATheBuilder
Copy link
Copy Markdown
Contributor Author

@3clyp50 Updates are done! I've moved the wraith evaluation prompt to a dedicated markdown file in a new prompts folder for easier review. Ready for your look.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants