HeartFlow AI Psychology Cognitive Engine
Paste the following prompt into your AI chat to install this skill:
Please follow https://skillhub.cn/install/skillhub.md to install @user_fdb96f95/heartflow-ai into your AI assistant.
About this skill
Problem
LLMs can lose user preferences, corrections, and task state across sessions, model changes, or restarts. Retrieval can be opaque, logic errors may surface with high confidence, and over-intervention can make agent responses feel noisy.
How It Works
HeartFlow separates memory, retrieval, verification, and governance into inspectable modules:
- MeaningfulMemory stores content in CORE, LEARNED, and EPHEMERAL layers
- HybridSearchEngine combines BM25, vector, and RRF fusion for retrieval
- SelfVerifier, CounterfactualEngine, and TruthfulnessChecker add inverse-consistency, counterfactual, and evidence-tracing checks
- EvolutionLoop records error strategies with Q-learning and favors repair paths with higher success rates
- PsychologyEngine and EmotionalProtocol calibrate emotional rationality and reduce unsupported certainty
It fits long-running agent work where persistent memory, traceable retrieval, and self-correction matter. It is not a chat persona pack. Confirm file permissions, tool access, and human review before deployment.
Use Cases
- Engineers maintaining a long-term knowledge base continue asking about preferences and past corrections after restarts.
- Developers building RAG apps retrieve hybrid results while preserving citations and evidence chains.
- Engineers debugging agents check logic chains, counterfactuals, and uncertainty before output.
- Architects governing multi-skill systems log error-code strategy success rates and roll back regressions.
Best For
- Engineers maintaining long-session agents need persistent preferences and correction logs across restarts.
- Developers building RAG retrieval services need BM25, vector search, citation tracing, and search transparency.
- Algorithm engineers debugging LLM output quality need counterfactuals, semantic entropy, and confidence calibration.
- Architects managing multi-agent skills need audit gates, self-healing, and version rollback.
Related Skills
A free matrix publishing assistant for Skill developers, guiding distribution to Tencent SkillHub, XiaLiao, XiaYou SkillHub, GitHub, and ClawHub with create, update, and delete safeguards.
A guide for writing Seedance 2.0 video prompts, covering @ references, camera moves, time segments, audio direction, and common pitfalls.
Collects daily conversations, task logs, and errors, then runs a nine-module review to surface blind spots, generate actions, and update memory.
Local long-term memory for AI agents that appends key facts and daily logs, supports cross-session context and search recall, and auto-archives old diaries to gzip files.