dsh-plugin-88api-image
88API Image Studio for DSH: six Image2, Image 2.5, and Nano Banana models for text-to-image, multi-reference editing, 2K/4K output, and sequential batches.
128 results
88API Image Studio for DSH: six Image2, Image 2.5, and Nano Banana models for text-to-image, multi-reference editing, 2K/4K output, and sequential batches.
DSH plugin: fixes compaction failure on local qwen3.8-27b gateways — xhigh thinking burns the entire output token budget, so thinking is off for compaction-only, with the model's non-thinking sampling parameters
DeepSeek Harness plugin that turns silent tool failures into loud ones: a tools/post-execute policy that matches warning signatures in successful tool output and blocks the result or attaches a notice
Model-visible redaction plugin for DeepSeek Harness: redacts secrets from model input and output streams
Interactive LaTeX copy affordances for DeepSeek Harness math output
Deterministic, CI-safe golden-output evaluation for DeepSeek Harness
Image generation for DeepSeek Harness via OpenRouter's unified Image API: text-to-image and image-to-image with capability-gated parameters, workspace file output, an inline chat card and a settings card
DSH web plugin: a draggable ring-gauge dashboard pinned to the bottom-right that shows live model output speed (tok/s), with an expandable detail panel (per-step exact speed, TTFT, cache-hit rate, cache read/write, reasoning and input tokens, model, cumulative tokens).
Turn-outline sidebar tab for dsh-better-sidebar: fold a session into user turns (input + tool steps + output) with structured jump-back into the conversation | DSH 轮次轨迹侧边栏:按用户轮次折叠会话(输入+工具步骤+输出),一键定位回对话原位
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。
Live streaming terminal output for DeepSeek Harness
DeepSeek Harness plugin converted from ayghri/i-have-adhd: ADHD-friendly output shaping that takes effect in a single session, never globally.
Open same-origin Markdown output links inside DSH Better Sidebar instead of an external app.
Provider discovery, matching, health checks, and pre-output failover for DeepSeek Harness.
dsh plugin: opencode-style json_schema structured output via a generic StructuredOutput tool, /json-schema command, and native JSON Schema validation.
Fix garbled non-ASCII output from the pwsh tool on Windows: native output (ANSI or UTF-8) is captured byte-preserving inside pwsh and decoded per-segment (strict UTF-8 else ANSI) before re-emitting as UTF-8. 修复 pwsh 工具中文/非 ASCII 输出乱码(兼容 GBK 与 UTF-8 两种原生输出)。
Dual-face (host + client) plugin: independent text-output ocr_image tool + tools/execute read_image capability dispatch + ctx.ocr seam, with a conversation.input.left OCR toggle button.
Always-on ADHD-friendly output for every DeepSeek Harness session: action first, numbered steps, task-progress restated. Host systemPrompt, not a skill catalog item.
DSH plugin: when a provider's model sync returns only ids, automatically fill display name, context window, max output tokens, reasoning efforts, and input modalities from the models.dev catalog (cached, TTL-refreshed), and surface a settings page with toggles, refresh, and a fill report.
DSH Web plugin: collapse each turn's tool calls into one group with a summary card, truncated output, and read-only filtering.
DSH web plugin: desktop notification, chime, and title flash when a session stops — finished output or waiting for human input. Click a notification to jump to that session.
DeepSeek Harness web GUI: collapse the agent trajectory (tool calls, thinking, workflow steps) while always keeping the agent's final output. A setting controls whether the trajectory is default-collapsed after the final output completes (default on).
S3-compatible spill backend for DeepSeek Harness: oversized tool output goes to object storage (AWS S3, MinIO, R2) instead of the agent's local disk
Indirect prompt-injection guard for DeepSeek Harness: taints tool output by origin and gates privileged tool calls that follow untrusted content