$ loading_
Generate metathoughts that help an agent inspect its own assumptions.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "io.github.ejentum/self-inspect-mcp" MCP server from askskill: Run: claude mcp add 'io-github-ejentum-self-inspect-mcp' -- npx -y self-inspect-mcp
Generate one metathought for this idea to help me inspect overlooked assumptions: 'We should prioritize this feature because users will definitely use it frequently.'
A concise reflection prompt highlighting possible assumptions about user need, usage frequency, or lack of evidence.
Given this thought, provide one metathought that makes the agent question its premises: 'The log error comes from the database, so restart the database first.'
A self-check prompt that can be inserted into an agent workflow to verify causality and alternative explanations.
Create one metathought for this statement to help me inspect my prior assumptions: 'The readers of this article definitely already understand the industry background.'
A reflective prompt reminding the user to validate audience knowledge level and whether more context is needed.
Provides reasoning, coding, anti-deception, and memory harness MCP tools.
Lets reasoning models delegate sub-tasks to extend analysis and execution.
Combine structured reasoning and execution so agents can think, act, and trace decisions.
Analyze complex problems with structured reasoning, self-critique, and iterative thinking.
Provide local-first semantic memory and context recall for MCP agents.
Extend model reasoning by delegating subproblems to a configurable auxiliary agent.