DeepSeek Harness PluginManifest Valid

DSH-Ling

A local-first memory, persona, and review loop plugin for DeepSeek Harness.

UI & ProductivityDeveloper Tools

1GitHub Stars0ForksUpdated2026-10-17

Source

Install

$ dsh plugin --profile web add @nightpainters/dsh-ling

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.

source-grounded

DSH-Ling (Artifact Spirit) is a local-first memory and persona management plugin for DeepSeek Harness. Built around a human-in-the-loop two-way review cycle, it structures agent memory into a verifiable three-layer hierarchy and an auditable memory tree. It features strict local persona locking, origin-checked administrative surgery gates, context retrieval tools (recall), and local SQLite-backed persistence without external telemetry.

Key Capabilities

Three-layer memory injection separating resident persona (L0), opening context summaries (L1), and active session runtime memories (L2)
Auditable memory tree with lineage-weighted retrieval (trunk, branches, hubs, and edges) that preserves content immutability
Persona freeze lock requiring local non-paste confirmation statements and host-level tickets for contractual changes
Multi-channel historical chat import supporting DSH sessions, web export databases, and standard JSON/JSONL archives
Recall tool providing progressive drill-down from conversation maps to exact past transcripts
Interactive dual modes (Work and Life) that adapt reasoning tiers, style blocks, and memory weighting vectors without switching models

Useful For

Maintaining persistent, evolving AI companion personas across lengthy projects without cloud data leakage
Structuring complex multi-session research and development workflows into categorized memory branches
Migrating and unifying fragmented historical conversation archives into a local DeepSeek Harness environment
Conducting deliberate periodic reviews of model habits, learned rules, and memory contradictions

Who It Fits

DeepSeek Harness power users demanding complete local data ownership and privacy
Developers seeking persistent, auditable AI collaboration companions for complex long-term projects
Researchers managing contextual knowledge trees and verifiable reference archives

Documented Limitations

Requires Node.js >=22.15.0 with built-in node:sqlite and zstd decompression support
Access log capture requires applying an optional core patch to the DSH profile
Persona surgery gates require direct local host loopback execution and cannot be modified remotely over proxies without disabling guards
Automatic deep summaries and opening memories follow scheduled batch intervals rather than instantaneous token-by-token updates

DSH Compatibility

Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.

Not tested yetNo runtime compatibility tests have been published yet.

Security Signals

Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.

Dependency Counts

package.json declares 1 runtime, 0 development, 13 peer, and 0 optional dependencies.

info

Dsh Bundle Declared

package.json declares DSH bundle metadata.

info

License Declared

GitHub reports the repository license as Apache-2.0.

info

Package Manifest Available

A root package.json was captured and can be inspected by the registry.

info

Source Available

Public GitHub source metadata is available for this registry snapshot.

info

Source & Registry Notes

Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.

Source repository: NightPainters/DSH-Ling
Registry source: GitHub · dsh-plugin topic
Registry classification: Plugin
Source checked: 64f47bd · 2026-10-19

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.