dsh-dupguard
Real-time repetition guard for DeepSeek Harness (DSH): stops model generation when the same string repeats >=10 times in the streamed output. 实时检测 DSH 大模型流式输出中的重复内容,同一字符串重复十次以上立即停止生成。
917 results
Real-time repetition guard for DeepSeek Harness (DSH): stops model generation when the same string repeats >=10 times in the streamed output. 实时检测 DSH 大模型流式输出中的重复内容,同一字符串重复十次以上立即停止生成。
On-demand vision for text-only LLMs in DeepSeek Harness: keep pasted/attached images in the backend, and let the model call view_image to look at them anytime via any OpenAI-compatible vision endpoint — as if the model were natively multimodal.
FriendliAI serverless LLM adapter for DeepSeek Harness — OpenAI-compatible chat completions with dynamic model discovery and reasoning support
Cron-based scheduled task plugin for DeepSeek Harness (dsh): pure UI management, zero LLM tools, cross-session timer, multi-level notifications.
cardian —— DeepSeek Harness 知识中心插件:RepoWiki / 知识卡片 / 记忆,全部落地到本地 Obsidian 仓库
Kiro CLI (ACP) adapter for the DeepSeek Harness LLM seam
WinAgent 的 DSH Web 插件:在 DSH 里使用 Windows 操作 Agent(OpenAI 兼容 API / 本地 Ollama,53+ Windows 工具,skills 与 MCP 挂载,LLM Wiki 个人知识库)
Google Antigravity / Cloud Code Assist model provider for DeepSeek Harness (DSH Desktop Beta Compatible)
Sync cc-switch provider profiles (Claude Code / Codex relays) into DeepSeek Harness as llm-pi-ai routes, credentials stored via the DSH credential seam.
dsh-langfuse: Langfuse LLM observability for DeepSeek Harness — one OpenTelemetry trace tree per agent session, with feedback scores and subagent lineage.
Cross-session persistent cost ledger for DeepSeek Harness: logs every LLM token usage to SQLite and exposes record/query/budget tools. Built-in DeepSeek pricing, overridable via config.
DSH Web GUI 模型能力配置:在「设置 → 模型」的模型目录编辑器内,为每个已配置的模型直接设置是否支持图像与思考等级(reasoning efforts),写入 llm-pi-ai 设置命名空间。
DSH harness bundle/plugin: tolerate OpenAI-compatible providers whose streaming responses omit finish_reason (e.g. the Snowflake Cortex gateway) so turns complete instead of failing with a TRANSPORT error.
Two-stage intent and evidence review, deterministic completion gates, and bounded repair for DeepSeek Harness agents
DeepSeek Harness plugin: use your own Claude and ChatGPT/Codex subscriptions as model providers, over each vendor OAuth flow
SQL 优化模式:EXPLAIN 分析、索引策略、N+1 解决、查询改写。受 wshobson/agents(38k★ MIT)启发。
Bridges vendor agent SDKs into DeepSeek Harness as selectable LLM providers. Works in any DSH mode; DSH keeps its own prompt, tools, sandbox and approvals.
A DeepSeek Harness (DSH) plugin that adds an AI-adjudicated 'smart' approval mode: safe operations are auto-allowed via an isolated side-channel LLM judge, everything else falls back to the normal human approval popup.
Explain failed DeepSeek Harness agent runs locally, then produce a redacted recovery plan. No model required.
第三方模型思考努力度配置(DeepSeek Harness 插件):在设置页为模型声明 reasoningEfforts 档位、路由默认档位与 Anthropic token 预算,写入 llm-pi-ai 命名空间;档位调节复用原生模型框 Effort 面板。
Agent JIT: a compiler/runtime that offloads deterministic execution paths from LLM agent loops.
Offline DeepSeek request, tool-loop, reasoning_content, and SSE diagnostics for DSH
DeepSeek Harness 插件:把文本形式的会话统计(轮/步、LLM/工具耗时、首 token、tok/s、缓存命中、输入/输出 token)替换为页面右侧直观醒目的图表面板,每秒刷新。
Gives a text-only LLM vision capability