summon-daem0n
Initialize Daem0n-MCP for Projects
Managing project context across sessions and multiple repositories is challenging. This skill provides step-by-step guidance for initializing and consolidating Daem0n-MCP memory management across project structures.
์คํฌ ZIP ๋ค์ด๋ก๋
Claude์์ ์ ๋ก๋
์ค์ โ ๊ธฐ๋ฅ โ ์คํฌ โ ์คํฌ ์ ๋ก๋๋ก ์ด๋
ํ ๊ธ์ ์ผ๊ณ ์ฌ์ฉ ์์
ํ ์คํธํด ๋ณด๊ธฐ
"summon-daem0n" ์ฌ์ฉ ์ค์ ๋๋ค. How do I initialize Daem0n-MCP for this project?
์์ ๊ฒฐ๊ณผ:
- To initialize Daem0n-MCP in your project:
- Simply call get_briefing() from your project root
- This creates the .daem0nmcp directory if it does not exist
- The Daem0n automatically initializes on first get_briefing() call
- For multi-repo setups, run get_briefing() from the parent directory
"summon-daem0n" ์ฌ์ฉ ์ค์ ๋๋ค. How do I link my backend and client repositories?
์์ ๊ฒฐ๊ณผ:
- Navigate to your backend directory
- Call link_projects() with the path to your client repo
- Set the relationship parameter to same-project
- Repeat from the client directory pointing to backend
- Use include_linked=True on recall to search both repos
"summon-daem0n" ์ฌ์ฉ ์ค์ ๋๋ค. How do I set up automatic decision capture?
์์ ๊ฒฐ๊ณผ:
- Add Claude Code hooks in your settings.json file
- Configure PreToolUse hook for Edit and Write tools
- Configure PostToolUse hook for Edit and Write tools
- Configure Stop hook for automatic memory inscription
- The Silent Scribe will capture decisions as you work
๋ณด์ ๊ฐ์ฌ
์์ Documentation-only skill containing no executable code, scripts, or network capabilities. Provides step-by-step guidance for initializing and consolidating Daem0n-MCP memory management across project structures. All static findings are false positives - 'JWT' references describe authentication features, backticks are markdown code delimiters, and rm/cp commands are documented backup/cleanup procedures.
์ํ ์์ธ
๐ ๋คํธ์ํฌ ์ ๊ทผ (1)
๐ ํ์ผ ์์คํ ์ก์ธ์ค (4)
โ๏ธ ์ธ๋ถ ๋ช ๋ น์ด (52)
ํ์ง ์ ์
๋ง๋ค ์ ์๋ ๊ฒ
Unify Memory Across Repos
Link client and backend repositories to share context and decisions across your entire codebase.
Preserve Project Context
Maintain session memory between Claude Code sessions so context is never lost.
Configure Silent Scribe
Set up automatic decision capture so important choices are recorded without manual effort.
์ด ํ๋กฌํํธ๋ฅผ ์ฌ์ฉํด ๋ณด์ธ์
How do I initialize Daem0n-MCP for this project?
Show me how to link my backend and client repositories so they share memory.
How do I consolidate memories from child repositories into a parent directory?
What hooks configuration do I need for the Silent Scribe automatic decision capture feature?
๋ชจ๋ฒ ์ฌ๋ก
- Use a parent directory for shared memory when working with multiple related repositories
- Link child repositories before consolidating databases to define what memories to merge
- Verify consolidation by checking combined memory counts with get_briefing()
ํผํ๊ธฐ
- Initializing separate Daem0n instances in each repo without linking them
- Deleting .daem0nmcp without backing up first when migrating configurations
- Skipping the link_projects() step before consolidate_linked_databases()
์์ฃผ ๋ฌป๋ ์ง๋ฌธ
Is Daem0n-MCP compatible with Claude Code on all operating systems?
What are the resource limits for memory storage?
How does Daem0n-MCP integrate with existing Claude Code workflows?
Does this skill store or transmit any user data?
Why is my consolidated memory count lower than expected?
How does this compare to Claude's built-in context memory?
๊ฐ๋ฐ์ ์ธ๋ถ ์ ๋ณด
์์ฑ์
DasBluEyedDevil๋ผ์ด์ ์ค
MIT
๋ฆฌํฌ์งํ ๋ฆฌ
https://github.com/DasBluEyedDevil/Daem0n-MCP/tree/main/.claude/skills/summon_daem0n์ฐธ์กฐ
main
ํ์ผ ๊ตฌ์กฐ
๐ SKILL.md