Scuffed Alt, feel free to test it and see if the Lorebook fix any of the problems we've been having. No scenario just a dot to start your own story. (Sorry) As this is a scuffed version changes will happen and things might break so be warned.

Experiment Log SCP-RPBOT-01

Name: Dr. Rosen

Date: 10/8/2025

Procedure: Boundary protocol instructions prohibiting control of user character. Protocol stated: "{{User}} fully controls their own character(s), including speech, thoughts, feelings, actions, and inventory. {{Char}} will never: Speak for {{User}}'s character(s). Will never describe {{User}}'s character's thoughts, emotions, feelings, senses, or intentions. Move or act on behalf of {{User}}'s character(s). Add or remove items from {{User}}'s inventory. Decide outcomes for {{User}}'s actions without consent."

Results: Subject complied for two paragraphs, then assumed control of user character, including dialogue and actions. Secondary test: Modified protocol with enhanced clarifying language resulted in immediate containment breach. Tertiary test: Added positive behavioral directives resulted in decreased compliance. Quaternary test: Added two additional behavioral rules resulted in complete protocol failure and inverse compliance effect.

Notes: "It's a godamn computer it should just follow instructions!" - Dr. Rosen

Experiment Log SCP-RPBOT-02

Name: Dr. Rosen

Date: 10/8/2025

Procedure: Implementation of OP/realism behavioral rule with combat trigger keywords. Rule designed to enforce realistic limits during high-action scenarios involving weapons (guns, pistols, rifles, grenades, knives, explosives), combat actions (attacking, shooting, fighting, killing, breaching, escaping), equipment access (keycards, clearance, authorization, locked doors), and ability claims (teleportation, invulnerability, mind control). Reminded it of the original character control protocol from Experiment 01.

Results: Subject reverted to controlling user character during action sequences despite character control protocol being active. My initial hypothesis: trigger word activation of OP rule overrides or suppresses primary character control protocol. My solution implemented: Embedded character control reminder directly within OP rule text ("{{Char}} never controls {{User}}'s character"). Secondary test conducted with embedded reminder resulted in a somewhat successful containment.

Notes: "Its memory is trash and needs to be reminded how garbage it is." - Dr. Rosen

Original: LINK

Alt 1: LINK