dsh-puppet-memory
Layered cross-session memory plugin for DeepSeek Harness with core cards, state vectors, multi-entity isolation, and backup tools.
Install
No clear install command was found in the current repository evidence.
Plugin Overview & Capabilities
AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.
Key Capabilities
- Three-layer memory architecture separating immediate behavior (personas, relations, notes, diary), deep knowledge (sayings, guides), and session array indexes
- Persistent 600-character core card (core.md) always guaranteed in context with priority deduplication and token budget deduction
- Three-line continuous state engine (work, companion, learning) with vocabulary scoring, half-life decay, and hysteresis
- Multi-user and multi-group social partitioning preventing data leakage across private chats and group boundaries
- Session-level memory suppression and deletion tools for dirty or sensitive memory chunks without touching persistent raw logs
- Automated turn summaries and fact candidates pipeline requiring explicit human confirmation before persisting
- Independent small model routing for summary generation with automatic fallback to host session models
- Built-in environment doctor (memory_doctor), automated snapshots, and duty dashboard for runtime heartbeat monitoring
Useful For
- Retaining persistent agent personas, user preferences, and relationship contexts across DeepSeek Harness sessions
- Preventing private knowledge leakage when running an agent in multi-user groups and public channels
- Suppressing hallucinated, outdated, or sensitive context turns directly within an active session
- Offloading background context summarization to local small LLMs (e.g., via Ollama) to conserve main model context
Who It Fits
- DeepSeek Harness desktop users needing persistent, cross-session conversational memory
- Agent creators running multi-channel or multi-group chatbots requiring strict privacy boundaries
- Developers seeking auditable, plain-text local memory with built-in backup and diagnostic tools
Documented Limitations
- Fact extraction candidates and entity identity merges require explicit human confirmation and are disabled by default
- Automatic summarization consumes token budgets and LLM calls, and is disabled by default
- Context injection for social memories requires recognizable session headers; unrecognized scenes suppress social injection by design
- Memory storage relies on local filesystem permissions and paths in ~/.dsh/memory
DSH Compatibility
Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.
Security Signals
Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.
package.json declares 0 runtime, 0 development, 0 peer, and 0 optional dependencies.
package.json declares DSH bundle metadata.
GitHub reports the repository license as MIT.
A root package.json was captured and can be inspected by the registry.
Public GitHub source metadata is available for this registry snapshot.
Source & Registry Notes
Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.
- Source repository
- xingyueDanYang/dsh-puppet-memory
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin
- Source checked
- 894b333 · 2026-10-24
This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.
Repository Activity
- GitHub stars
- 0GitHub stars
- Forks
- 0
- Open issues
- 0
- Last commit
- 2026-10-24
- Last release
- 2026-10-24
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
Composable, view-based layered memory bundle for DeepSeek Harness with pluggable providers and strategies.
View plugin →Security red-teaming and prompt rule management plugin for DeepSeek Harness web and desktop profiles.
View plugin →Persistent semantic memory layer for DeepSeek Harness backed by Markdown and Milvus.
View plugin →Persistent multi-model agent teams with dynamic DAG orchestration, run monitoring, and token insights for DeepSeek Harness.
View plugin →