autodream-core

v1.0.0

通用记忆整理引擎 — 基于适配器模式的跨平台记忆整理技能。自动去重、合并、删除过时条目。| Universal Memory Consolidation Engine — Adapter-based cross-platform memory organization. Auto-dedup, merge, pr...

0· 58·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
medium confidence
Purpose & Capability
The name/description (memory consolidation for agents) aligns with the code: adapters for a workspace, stages to scan memory files and sessions, dedupe/merge/prune logic, and an OpenClaw adapter. All functionality (file reads/writes, session scanning, index updates) is explained and needed for the stated purpose.
Instruction Scope
SKILL.md and runtime code instruct the agent to scan the workspace (MEMORY.md, memory/*.md, sessions/*.json), extract entries, deduplicate, prune stale entries, and update MEMORY.md. It also writes local state (.dream_state.json) and analytics (.autodream_analytics.jsonl). The instructions do not attempt to read unrelated system config, environment variables, or send data externally.
Install Mechanism
Registry has no formal install spec (instruction-only), which minimizes automatic install risk. However the package includes install.sh and package.json in the bundle — these are not executed by the platform automatically but could be used by a user to install. Inspect install.sh/package.json before running them; their presence is a packaging inconsistency but not proof of maliciousness.
Credentials
The skill requires no environment variables, no external credentials, and uses only local filesystem access in the configured workspace. The analytics logger writes local JSONL files; no network/external endpoints are used. Credential access requests are proportionate (none requested).
Persistence & Privilege
The skill persists state and analytics under adapter.memory_dir/autodream (creates directories and writes .dream_state.json and .autodream_analytics.jsonl). always is false and model invocation is allowed (platform default). Persisting local state is expected for this tool but you should be aware it will create/modify files in the specified workspace.
Assessment
This package appears coherent for a local memory-consolidation library: it reads your workspace (MEMORY.md, memory/*.md, sessions/*.json), creates state and analytics files under the workspace, and does local-only processing. Before installing or running: 1) Inspect install.sh and package.json before executing any install script. 2) Run first in a non-production/test workspace (set adapter.workspace to a safe folder) and use engine.run(force=True) manually per the README. 3) Backup your MEMORY.md (as the docs advise). 4) If you don't want local analytics, disable enable_analytics in the config or delete/ignore .autodream_analytics.jsonl. 5) Note minor packaging inconsistencies (package.json and NPM links vs. a Python implementation) — this is likely a metadata artifact but worth confirming the author's source (GitHub link in metadata points to a placeholder). If you want higher assurance, ask the publisher for provenance (repo URL and install instructions) and/or perform a quick code review of install.sh before use.

Like a lobster shell, security has layers — review code before you run it.

latestvk9755bdv1cz6422dd65295dp8n845pw9

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments