Track agent capabilities, knowledge, confidence, and cognitive state via CLI.
npx skills add https://github.com/Heretek-AI/heretek-openclaw --skill self-model
Please help me install this Agent Skill. Skill: self-model Source: https://github.com/Heretek-AI/heretek-openclaw/tree/main/skills/self-model Command: npx skills add https://github.com/Heretek-AI/heretek-openclaw --skill self-model
Meta-cognitive self-awareness for agent collectives, enabling understanding of their own capabilities, knowledge, and cognitive state to improve reliability and learning.
Run node self-model.js --summary to view the agent's current self-model summary.
High-intent search queries and answers about installing and using this skill.
Build metacognitive self-awareness by implementing a persisted self-model that tracks agent capabilities, knowledge domains, confidence levels, and cognitive state to improve collective reliability and learning progression.
Agent capability tracking is the registration, activation, and learning of new capabilities to adapt to tasks. It helps autonomous workflows by allowing agents to understand their own boundaries and adapt within OpenClaw environments.
Track AI agent cognitive state and confidence levels by maintaining a persisted self-model that organizes knowledge by domain, monitors focus and thinking state, and surfaces confidence trends accessible via CLI or programmatic API.
Yes, you can monitor active tasks and self-reflection across multiple domains using a self-model that supports task tracking, cognitive state management, and reflection capabilities for autonomous agent collectives.
View the agent's current capability and knowledge summary from the command line by running the self-model summary command, which interrogates the persisted self-model to display registered capabilities, cognitive state, and metrics.
Limitations of a persisted self-model include the overhead of continuously tracking cognitive state and confidence levels, which requires consistent capability registration and knowledge management to maintain accurate self-reflection across agent collectives.
1molchuanVerify MCP, Codex, and Gemini endpoints and initialize RPI state.
DojoGenesisBreak learning goals into chunked study sessions with comprehension gates.
FlexNetOSAdd datestamped progress notes to task entries in a knowledge base system.
vvieira010-pixelCompare learner confidence ratings with performance to identify calibration mismatches.
