Retrieve personal wiki pages via keyword and semantic search with citations.
npx skills add https://github.com/vietbui1999ru/dotfiles --skill wiki-vietbui1999ru
Please help me install this Agent Skill. Skill: wiki Source: https://github.com/vietbui1999ru/dotfiles/tree/main/llm-wiki-plugin/skills/wiki Command: npx skills add https://github.com/vietbui1999ru/dotfiles --skill wiki-vietbui1999ru
It reduces time spent re-deriving agent engineering and security best practices by quickly retrieving relevant notes from your personal LLM wiki.
Use case: You are designing an agent plugin for context management and need known guidance on context compression, indirect prompt injection defenses, and verification pipelines—use this Skill to load those pages before proposing an architecture.
Use the wiki skill to search for relevant concepts for your current task and load the best-matching pages into context before you start designing or answering.
High-intent search queries and answers about installing and using this skill.
Semantic wiki retrieval extracts task-specific keywords and searches your personal knowledge base to surface relevant LLM patterns. It loads up to three high-scoring concept pages for agent engineering and architectural decisions into your active context.
Use keyword and semantic wiki retrieval to search your personal LLM knowledge base for high-signal concept pages. The system prioritizes loading full concept pages over summaries to deliver better downstream answers for security hardening and context compression strategies.
Yes, wiki retrieval supports scenarios like designing agent orchestration, managing context windows, hardening against indirect prompt injection, and selecting retrieval and verification strategies. It extracts two to three task-specific terms to find the most relevant pages for your architecture.
Source citation enforcement requires citing the loaded wiki pages when applying patterns to your agent engineering tasks. It also flags newly discovered reusable patterns as wiki candidates, ensuring your personal knowledge base continuously captures emerging architectural decisions.
Retrieval is limited to extracting two to three task-specific terms and loading a maximum of three pages above a minimum score threshold. Results depend on the existing content in your personal wiki and prioritize concept pages over general summaries.
gustavofsantosIndex local Markdown files and search them by keyword or natural language.
Jony2176-cloudBuild AI agents and multi-agent systems with the Agno framework.
vietbui1999ruSearch the personal LLM wiki and cite up to three relevant pages.
andrewblockernstImplements vector-based semantic search for documents using AgentDB and Node.js.
