dsh-darwin
Dual-plugin self-evolution architecture for DeepSeek Harness combining session issue mining, tiered candidate synthesis, and rollback.
Install
$ dsh plugin --profile <你的profile> add dsh-darwin-sentinelPlugin 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
- Scans recent DSH session logs with darwin_scan to mechanically identify retry loops, tool error clusters, and token waste
- Generates structured health and diagnostics reports via darwin_report with issue deduplication and fingerprinting
- Synthesizes zero-code candidate skills and config patches via darwin_forge based on high-severity tickets
- Promotes verified candidates to project .dsh/skills with hot-reload and mandatory snapshot creation
- Performs deterministic skill rollback using stored snapshots via darwin_rollback
- Isolates regression test evaluation criteria (redactForAgent) to prevent reward hacking during evolution
Useful For
- Analyzing repetitive agent failures or tool execution loops across historical DSH sessions
- Generating automated Markdown-based skill definitions to address discovered failure modes
- Safely evaluating and promoting self-generated workflow improvements with rollback guarantees
Who It Fits
- DeepSeek Harness plugin developers and power users building autonomous feedback loops
- Engineers seeking automated session diagnostics and failure pattern detection in DSH
Documented Limitations
- Code-level synthesis (Creator mode) is disabled by default for security, supporting only skill and config synthesis in current versions
- Requires Node >= 22 and targets DeepSeek Harness 0.1.1-rc.2 through 0.1.2-alpha.x preview interfaces
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.
Source & Registry Notes
Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.
- Source repository
- que3sui/dsh-darwin
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin Bundle
- Source checked
- a1abdaf · 2026-09-17
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
- 1GitHub stars
- Forks
- 0
- Open issues
- 2
- Last commit
- 2026-09-14
- Last release
- 2026-09-02
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
Comprehensive long-term memory, self-evolution, skills, task management, and multi-agent dispatch bundle for DeepSeek Harness.
View plugin →Previewable cross-preset session migration for DeepSeek Harness with bounded, fixed-schema handoffs.
View plugin →Persistent multi-model Agent teams and bounded DAG workflows with GUI controls for DeepSeek Harness.
View plugin →Second-model AI auto-review for DeepSeek Harness approval requests with fail-closed safety and session audit.
View plugin →