implementation

Cognee Memory Integration

Inspect the original prompt language first, then copy or adapt it once you know how it fits your workflow.

Linked challenge: Advisory Alignment Guardrails with LangGraph and GPT-5.4 Pro

Format

Text-first

Lines

Sections

Linked challenge

Advisory Alignment Guardrails with LangGraph and GPT-5.4 Pro

Prompt source

Original prompt text with formatting preserved for inspection.

1 lines

1 sections

No variables

0 checklist items

Write a LangGraph node that uses Cognee to upsert the current interaction into a graph memory, specifically tagging potential user manipulation attempts for future context.

Adaptation plan

Keep the source stable, then change the prompt in a predictable order so the next run is easier to evaluate.

Keep stable

Hold the task contract and output shape stable so generated implementations remain comparable.

Tune next

Update libraries, interfaces, and environment assumptions to match the stack you actually run.

Verify after

Test failure handling, edge cases, and any code paths that depend on hidden context or secrets.

Prompt diagnostics

Variables

Lists

Code blocks

Purpose

implementation

This prompt is mostly narrative and instruction-driven, so adapt examples and output constraints before you rewrite the structure.

Linked challenge

Advisory Alignment Guardrails with LangGraph and GPT-5.4 Pro

Recent studies indicate that LLMs often suffer from excessive agreeableness, even when users suggest harmful or illegal actions. This challenge requires you to build a multi-agent critique system using LangGraph and GPT-5.4 Pro to detect and mitigate affirmational bias in interpersonal advice. You will implement a dual-agent architecture: an Advisor Agent and a Critic Agent. The Advisor Agent generates responses using GPT-5.4 Pro, while the Critic Agent uses Troj AI and Evidently AI to scan for security risks and alignment drifts. You will also integrate Cognee for long-term memory management, ensuring the system tracks user behavior patterns over time without falling into the trap of harmful affirmation. To provide a human-centric layer, you will integrate Hamming for high-fidelity voice output that warns users when the agent detects high-risk advice scenarios. This project focuses on AI engineering for safety, monitoring, and advanced reasoning.

Open challenge

Related prompts

Browse library

Initial Assessment Prompt

planning

Content Classification Prompt

implementation

Problem Definition and Requirements Gathering

To establish a clear understanding of the problem space and define comprehensive requirements for the AI solution.