Skip to content
AI Agents

Agent Memory Viewer and Editor

See, search, correct and delete what your agent remembers about a user.

Intermediate16h1–2 people15/20 fit
TL;DR

Build Agent Memory Viewer and Editor for developers building assistants with memory in ~16h with 1–2 people. Stack: Vector DB + LLM + Next.js.

Why it wins: Transparency and user control are strong judging points.

Build flow

01 · Problem
Agent memory is a black box, so wrong or sensitive memories persist silently.
02 · Solution
Store memories as structured records with sources, show them in a dashboard, let users edit or forget items and test how answers change.
03 · Build
  1. 1Extract memories from chats
  2. 2Store them with embeddings and sources
  3. 3Build the dashboard
  4. 4Re-run queries after edits
04 · Demo
Delete a wrong memory and show the agent's answer change immediately.

MVP checklist

  • Memory timeline
  • Source for each memory
  • Edit and forget
  • Before and after test

Tech stack

TechRole
Vector DBEmbedding search
LLMLanguage model
Next.jsFrontend + API routes

Reference repos

Open-source projects you can build on. Links checked 2026-10-31.

Related ideas

More AI Agents

  • AI Agents

    Code Review Agent

    Catches risky changes before a human reviewer looks at the PR

    Intermediate36h2–3
    PythonGitHub APILLM
    16/20

  • AI Agents

    Agent Trajectory Evaluator

    Replay agent runs, score each step and find where agents go off the rails

    Advanced26h1–3
    OpenTelemetryLLMNext.js
    15/20