dsh-memory
Qoder-Agent-SDK-style persistent dual-scope memory with budgeted context injection for DeepSeek Harness.
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
- Dual-scope memory management supporting global user-level and project-specific knowledge bases under $DSH_HOME
- Separated generation (background write pass) and consumption (token-budgeted context injection) lifecycles
- Delta-only dynamic projection during agent steps to minimize redundant context re-injection
- Just-in-time (JIT) path matching to dynamically load project memory based on files touched in session
- Daily memory journal logging to capture high-value turn outcomes in non-polluting subdirectories
- Comprehensive Web UI settings panel under Settings -> Memory with injection preview and scope management
- Dedicated memory inspection commands (/memory, /memory-switch, /memory-budget, /memory-trust, etc.)
- Fine-grained sandbox-aware filesystem interactions respecting Harness write policies
Useful For
- Retaining cross-session project architectural decisions, conventions, and context without checking files into Git
- Maintaining global user preferences, workflows, and behavioral guidelines across all Harness sessions
- Selectively loading project-specific rules only when specific file globs or modules are being touched
- Logging turn-by-turn engineering breakthroughs and daily operational progress via the memory journal
Who It Fits
- DeepSeek Harness developers seeking persistent memory across CLI and web chat sessions
- Engineers working on complex codebases requiring persistent architectural context without workspace pollution
- Users migrating from or accustomed to Qoder Agent SDK memory conventions
Documented Limitations
- Project key identifiers inherit Harness session-persistence lossy normalization (dashes and path separators may collide)
- Harness file system APIs lack file modification timestamps (mtime) and system file explorer integrations
- Daily journals are deliberately omitted from memory list/read/search tools to avoid context inflation
- Updates performed via git fast-forward only take effect after restarting the dsh web server host process
- Depends on @deepseek-ai/schemastery which must be present in the installed Harness environment
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 2 runtime, 3 development, 0 peer, and 0 optional dependencies.
package.json declares DSH bundle metadata.
GitHub did not report an SPDX license for this repository snapshot.
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
- hongweifei/dsh-memory
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin
- Source checked
- bd09744 · 2026-10-20
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
- 7GitHub stars
- Forks
- 0
- Open issues
- 0
- Last commit
- 2026-10-16
- Last release
- No release detected
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 →LLM-supervised persistent memory and graph-based recall integration bundle for DeepSeek Harness.
View plugin →Second-model AI auto-review for DeepSeek Harness approval requests with fail-closed security.
View plugin →Lingshu long-term auditable memory system and cognitive graph for DeepSeek Harness.
View plugin →