Zum Hauptinhalt springen
Dieser Inhalt ist noch nicht in Ihrer Sprache verfügbar und wird auf Englisch angezeigt.

Hindsight Local

Marketplace Aktiv

Hindsight: Agent Memory That Learns

1 Plugin
Zweck

To provide AI agents with a sophisticated memory system that facilitates learning and long-term recall, enhancing their ability to handle complex tasks and approximate human-like intelligence.

Funktionen

  • Learns over time, not just recalls history
  • Biomimetic memory organization (World, Experiences, Mental Models)
  • High-performance memory recall and reflection capabilities
  • Multiple integration methods (SDKs, Docker, embedded)

Anwendungsfälle

  • Personalizing AI chatbots with per-user memories
  • Enabling AI employees to handle open-ended tasks and adapt behavior
  • Automating complex tasks by learning from experiences

Nicht-Ziele

  • Simple conversation history recall as the primary function
  • Basic RAG or knowledge graph limitations
  • Overkill for simple AI workflows with basic needs

Documentation

  • info:Configuration & parameter referenceWhile the README provides installation and basic usage examples, a detailed reference for all configuration parameters and their defaults across different LLM providers and storage options is not explicitly documented.

Security

  • info:Secret ManagementThe documentation mentions setting API keys (e.g., `OPENAI_API_KEY`) via environment variables for LLM providers, but there is no indication of secrets being echoed to stdout or committed.

Trust

  • warning:Issues AttentionThe repository has 70 open issues and 233 closed issues in the last 90 days, resulting in a closure rate of approximately 77% (233 / (70 + 233)), which is good. However, the high number of open issues (70) compared to the closure rate suggests maintainer bandwidth might be stretched, warranting a warning for potential slowness in addressing all issues.

Compliance

  • info:GDPRThe extension handles user-provided content for memory storage and recall. While it doesn't explicitly claim to process personal data, its nature suggests personal data might be involved, and there's no explicit mention of GDPR-specific sanitization beyond what the LLM might naturally do.
  • info:Telemetry opt-inThe README mentions a telemetry pixel in the footer for usage tracking, but it's unclear if this is opt-in or opt-out, or if it collects any specific data beyond basic analytics.

Installation

/plugin marketplace add vectorize-io/hindsight

Qualitätspunktzahl

88 /100
Analysiert about 14 hours ago

Vertrauenssignale

Letzter Commitabout 15 hours ago
Sterne13.2k
LizenzMIT
Status
Quellcode ansehen