Skillget

bamdra-openclaw-memory

agent
Guvenlik Denetimi
Basarisiz
Health Uyari
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 7 GitHub stars
Code Basarisiz
  • rm -rf — Recursive force deletion command in package.json
  • process.env — Environment variable access in packages/bamdra-memory-context-engine/src/index.ts
Permissions Gecti
  • Permissions — No dangerous permissions requested
Purpose
This agent provides a durable memory and context management runtime for OpenClaw sessions. It helps maintain conversation continuity and creates a searchable knowledge base from local markdown documents.

Security Assessment
The overall risk is rated as Medium. The tool accesses environment variables, which is typical for configuring memory and storage paths, but requires caution to ensure no sensitive credentials are leaked.

The most significant concern is a failed audit check for a recursive force deletion command (`rm -rf`) located in the `package.json` file. While this might simply be a cleanup script used during installation or updating, force deletion commands can be dangerous if they mishandle file paths. The tool does not request explicitly dangerous system permissions, and no hardcoded secrets were detected. Given its deep integration with local user files and directories, you should verify exactly what the deletion script targets before running it.

Quality Assessment
The project is actively maintained, with its most recent push occurring today. It uses the standard, permissive MIT license. However, it suffers from low community visibility, having only 7 stars on GitHub. This means it has not been widely vetted by the open-source community, increasing the risk of undetected bugs or vulnerabilities.

Verdict
Use with caution — manually verify the `rm -rf` script in `package.json` before installing, and be aware that the tool has very low community oversight.
SUMMARY

Give one OpenClaw session durable memory, topic-aware continuity, and bounded token growth.

README.md

bamdra-openclaw-memory

Bamdra Animated Logo

The continuity-first memory runtime for OpenClaw.

Install once:

openclaw plugins install @bamdra/bamdra-openclaw-memory

Upgrade later with the same command. If you need to pin a release, add the version suffix, for example @bamdra/[email protected].

That single install now prepares the full Bamdra memory stack:

  • bamdra-openclaw-memory
  • bamdra-user-bind
  • bamdra-memory-vector

If you already have an older or broken local install, use the standalone Clawdhub repair skill:

clawdhub --workdir ~/.openclaw --dir skills install bamdra-memory-upgrade-operator --force

中文文档

What it is

bamdra-openclaw-memory is the main runtime plugin in the Bamdra suite.

It helps OpenClaw:

  • keep the right topic branch alive
  • store durable facts
  • assemble compact prompt context
  • recover after interruptions and restarts
  • work with a real identity layer and a real local knowledge base

Why teams install it

Without a continuity runtime, long OpenClaw sessions break down fast:

  • users repeat the same background
  • interruptions destroy flow
  • stable decisions sink into chat history
  • local docs and notes are not part of the recall path

With this suite, the assistant can gradually evolve with the user because profile, memory, and knowledge all become durable.

What makes the suite complete

bamdra-user-bind

Adds the stable user boundary and living profile layer.

It covers most of what a per-user USER.md would usually try to do:

  • preferred address
  • timezone
  • tone preferences
  • role and collaboration style
  • long-lived user notes

Recent profile behavior improvements:

  • profile updates now support semantic replace, append, and remove instead of only blind overwrite
  • the Markdown mirror keeps machine-readable frontmatter as the controlled source
  • the Confirmed Profile section in the body is now a human-readable mirror of the same structured fields

bamdra-memory-vector

Turns local Markdown into a real knowledge base.

It indexes:

  • knowledge/
  • docs/
  • notes/
  • ideas/

and makes local recall happen before unnecessary web lookup.

Architecture

Bamdra Suite Architecture

Best-practice setup

Use the suite like this:

  • let bamdra-openclaw-memory own the memory and contextEngine slots
  • let bamdra-user-bind own identity and personalization
  • let bamdra-memory-vector own local Markdown knowledge and semantic recall

Recommended vector roots:

{
  "plugins": {
    "entries": {
      "bamdra-memory-vector": {
        "enabled": true,
        "config": {
          "enabled": true,
          "privateMarkdownRoot": "~/Documents/Obsidian/MyVault/openclaw/private",
          "sharedMarkdownRoot": "~/Documents/Obsidian/MyVault/openclaw/shared",
          "indexPath": "~/.openclaw/memory/vector/index.json"
        }
      }
    }
  }
}

That gives you:

  • a local-first memory runtime
  • a living user profile
  • a maintainable private and shared knowledge base

Profile update semantics

bamdra-user-bind no longer treats every profile change as a full replacement.

When the user says something new about how they prefer to collaborate, the runtime now distinguishes between:

  • replacing an old preference
  • appending another stable preference
  • removing one specific old preference

This matters for fields like preferences, personality, and notes, where the right behavior is often incremental rather than destructive.

The Markdown mirror is also clearer now:

  • frontmatter remains the machine-readable source of truth
  • the body shows a synchronized human-readable summary
  • Supplementary Notes are reserved for durable context that does not fit the structured fields

Repository map

Yorumlar (0)

Sonuc bulunamadi