dsh-memory
Cross-session plaintext memory plugin for DeepSeek Harness with deterministic BM25 recall and budgeted prompt injection.
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
- Two-tier plaintext JSON storage partitioned into global (~/.dsh) and project-level (<cwd>/.dsh) namespaces
- Deterministic BM25 text retrieval with CJK 2-gram tokenization and optional pluggable semantic embedding fusion
- Hierarchical workspace awareness allowing subdirectories to query upward along the ancestor chain
- Budgeted system prompt recall injection with summary truncation and overflow counters
- Validity anchors linking memories to environment variables or tool states to flag stale memories automatically
- Complete suite of 12 tools including memory management (save, list, search, update, move, forget, sweep) and prompt template utilities
- Multi-instance conflict avoidance using pre-write fingerprint checks (mtime and size comparison)
Useful For
- Persisting project-specific technical conventions and architectural decisions directly in Git-tracked repository files
- Retaining cross-session user preferences and personal guidelines across multiple DSH sessions globally
- Organizing and searching shared prompt template libraries alongside session memory
- Safely capturing and recalling decisions with complete auditability and manual overrides in settings UI
Who It Fits
- DeepSeek Harness users needing persistent, auditable cross-session memory without vector database overhead
- Development teams collaborating on shared repository context via Git-committed project memory
- SSID desktop ecosystem users
Documented Limitations
- Concurrent writes within the exact same millisecond window across separate instances resolve to last-write-wins
- workspace.json registration is handled by the upstream DSH core and remains outside the plugin's conflict-mitigation boundary
- Semantic embedding fusion requires an external embedding provider implementation configured in cordis.yml
- Requires DeepSeek Harness host environment providing @deepseek-ai/cordis and peer DSH storage/tool packages
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 1 runtime, 12 development, 7 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
- Max-Null/dsh-memory
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin
- Source checked
- df3cbe9 · 2026-10-16
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
- 5GitHub stars
- Forks
- 0
- Open issues
- 0
- Last commit
- 2026-10-02
- Last release
- 2026-09-28
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
Lingshu long-term auditable memory system and cognitive graph for DeepSeek Harness.
View plugin →LLM-supervised persistent memory and graph-based recall integration bundle for DeepSeek Harness.
View plugin →Persistent semantic memory layer for DeepSeek Harness backed by Markdown and Milvus.
View plugin →Composable, view-based layered memory bundle for DeepSeek Harness with pluggable providers and strategies.
View plugin →