DeepSeek Harness PluginIndexed

dsh-evolution

Self-evolving memory and skill management plugin family for DeepSeek Harness.

Developer ToolsSecurity & PolicySkills & Workflows

5GitHub Stars2ForksUpdated2026-10-24

Source

Install

$ dsh plugin --profile web add @lmzhen/dsh-evolution-all # the DEFAULT full bundle

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-evolution is a suite of 30 composable Cordis plugins designed for DeepSeek Harness. Inspired by Hermes Agent, it equips agents with self-evolution capabilities through automated session reviews, persistent long-term memory updates, and dynamic skill generation, patching, and curation with built-in threat scanning and version rollback controls.

Key Capabilities

Automatic session reviews and memory extraction into persistent local Markdown files
Dynamic skill generation, patching, and cataloging via model tools
Deterministic skill curation, duplicate merging, and archiving
Version rollback supporting up to 20 historical versions per skill
Security gating with threat scanning and optional human approval workflow (M2 mode)
GUI integration for parameter configuration and skill history visualization
Slash commands including /evolution doctor, /evolution pending, /evolution curator, and /graph

Useful For

Retaining project context and domain knowledge across fresh DSH sessions
Enabling agents to continuously learn and optimize their own skill definitions from task outcomes
Reviewing and approving agent-proposed system changes and skill updates
Managing and pruning expanding collections of agent skills automatically

Who It Fits

Developers looking for cross-session agent memory and adaptive capabilities
Power users managing complex workflows and custom skill sets in DeepSeek Harness
Teams seeking local, auditable agent evolution without third-party hosted telemetry

Documented Limitations

Default inject review mode processes reviews in-session without generating a separate replay activity ledger
Tightly coupled with specific DSH platform dependency lines (^0.1.5-rc.2)
Host-only mode (M3) disables direct model tool access for skill generation and memory manipulation
Requires Node.js >= 22.19.0 or >= 24.0.0

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 0 runtime, 0 development, 0 peer, and 0 optional dependencies.

info

License Declared

GitHub reports the repository license as MIT.

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: lmzhen/dsh-evolution
Registry source: GitHub · dsh-plugin topic
Registry classification: Plugin Bundle
Source checked: 862a22f · 2026-10-27

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.