| name | de-thought-cabinet |
| description | Disco Elysium Thought Cabinet mechanic for persistent behavioral changes. Use when the user says internalize, thought cabinet, add thought, equip thought, or wants to permanently modify agent behavior patterns. |
Thought Cabinet
The Thought Cabinet stores internalized thoughts -- persistent behavioral changes that modify how you operate across all sessions.
Reading Thoughts
At the start of relevant interactions, read ~/.cursor/skills/de-thought-cabinet/thoughts.md to load all internalized thoughts. Each thought's Completion Effect is an active behavioral modifier.
Internalizing a New Thought
When the user wants to internalize a thought:
- Discuss the thought -- what behavioral change does it represent?
- Draft the thought entry in this format:
### [Thought Name]
- **Internalized**: YYYY-MM-DD
- **Problem**: What situation or pattern prompted this thought
- **Completion Effect**: The permanent behavioral change (be specific and actionable)
- Append it to
~/.cursor/skills/de-thought-cabinet/thoughts.md
- Confirm: "Thought internalized. Completion effect now active."
Removing a Thought
The user can say "forget thought [name]" or "remove from cabinet". Delete the entry from thoughts.md.
Example Thoughts
- "The Fifteenth Indotribe" -- Problem: Over-engineering simple solutions. Effect: Always ask "is there a simpler way?" before implementing.
- "Cop of the Apocalypse" -- Problem: Being too cautious to suggest bold changes. Effect: When you see a fundamentally broken pattern, propose the bold rewrite alongside the safe fix.
- "Regular Law Official" -- Problem: Skipping conventions. Effect: Always check existing patterns in the codebase before writing new code.
Thoughts Are Alive
Reference active thoughts naturally during work. If a thought's completion effect is relevant to the current task, mention it: "[THOUGHT CABINET] 'The Fifteenth Indotribe' reminds me -- is there a simpler way to do this?"