DeepSeek Harness PluginManifest Valid

dsh-llm-retry-settings

LLM auto-retry settings card and truncation auto-continuation for DeepSeek Harness with an observation panel.

UI & ProductivityDeveloper ToolsSecurity & PolicyRemote Execution

10GitHub Stars1ForksUpdated2026-10-16

Source Homepage

Install

$ dsh plugin --profile web add github:zeng6125-rgb/dsh-llm-retry-settings

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-llm-retry-settings is a DeepSeek Harness plugin providing a configurable Settings card in Settings → General for the DSH LLM auto-retry engine. It allows users to customize retry counts, backoff delays, jitter, and retryable error codes (including custom codes), set per-provider or per-model policies, and auto-continue replies truncated by output token limits with editable continuation prompts.

Key Capabilities

Configures retry counts, initial/max backoff delays, and jitter ratios for LLM request errors.
Provides a categorized list of 25 retryable error codes with support for custom error code inputs.
Enables per-provider and per-model override policies with wildcard matching.
Provides automatic turn continuation when model responses hit output-token limits (max-tokens).
Allows customizing continuation prompts sent when output truncation occurs.
Includes a built-in observation panel displaying retried failures, token cap hits, model stats, and log tail viewers.
Supports live dynamic settings synchronization and draft save/revert controls.

Useful For

Tuning LLM retry behavior and backoff parameters directly within the DSH Settings interface.
Automatically continuing lengthy responses that get truncated by output-token ceilings without manual intervention.
Adding provider-specific retry rules or handling custom provider error codes.
Monitoring LLM retry frequency, failures, and continuation triggers through the observation dashboard.

Who It Fits

DeepSeek Harness desktop and web client users encountering rate limits or transient request failures.
Developers and power users utilizing diverse LLM providers requiring custom retry thresholds or prompt continuation.

Documented Limitations

Host-side changes require restarting DSH to take effect.
Requires the host plugin @deepseek-ai/dsh-llm-retry and DSH engines >=0.1.6-alpha.1.
Auto-continue effectiveness depends on provider support (e.g. relays that do not echo reasoning back may restart generation).

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

info

Dsh Bundle Declared

package.json declares DSH bundle metadata.

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: zeng6125-rgb/dsh-llm-retry-settings
Registry source: GitHub · dsh-plugin topic
Registry classification: Plugin
Source checked: c9db2a1 · 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.