Develop and extend AI agent components with vault-first knowledge retrieval.
npx skills add https://github.com/adrozdenko/soleri --skill agent-dev
Please help me install this Agent Skill. Skill: agent-dev Source: https://github.com/adrozdenko/soleri/tree/main/agents/salvador-filetree/skills/agent-dev Command: npx skills add https://github.com/adrozdenko/soleri --skill agent-dev
This Skill streamlines the process of developing and extending the AI agent's own codebase, ensuring that all internal changes are made with a deep understanding of existing architecture and best practices.
Use the agent-dev skill to add a new facade to the agent.
High-intent search queries and answers about installing and using this skill.
To extend an AI agent's internal tools while maintaining architecture, enforce a vault-first knowledge retrieval strategy that searches for existing patterns and validates them before modifying any code.
Vault-first knowledge capture in agent development prioritizes querying the agent's internal knowledge base to retrieve established architecture and proven brain patterns before inspecting code or implementing new features.
To add a new facade to an AI agent codebase, follow a structured development workflow that searches the vault for existing tool patterns, validates brain patterns, and then reads the relevant code to ensure consistency.
Yes, updating the vault after modifying agent brain features is required to capture new learnings back into the knowledge base, ensuring future development cycles can leverage the newly acquired intelligence.
General knowledge is not the primary source for agent codebase extensions; the workflow mandates explicit vault searches and pattern validation prior to implementation to ensure strict adherence to existing architecture.
Explicit vault searches are required in AI agent development to enforce a vault-first retrieval strategy, guaranteeing that new facades, tools, and brain features align with proven architectural patterns before code modification.
oaustegardCoordinate multiple AI agents to analyze code concurrently from security, performance, and architecture perspectives.
Smith-HappensValidate implemented products against PRD requirements extracted from documents.
haposoftTransforms raw ideas into validated, spec-ready designs for software architecture.
DayzideeAudit a repository for over-engineered code and suggest standard library replacements.
