dsh-turn-fork
Turn-atomic message editing, retry, reroll and version-tree navigation for DeepSeek Harness conversations
395 results
Turn-atomic message editing, retry, reroll and version-tree navigation for DeepSeek Harness conversations
Tacit learns what you leave unsaid in your prompts — from messy turns and your own corrections, with zero clicks — and tells the agent how to compensate, on every turn, via a system-prompt section you can read and edit.
Memory protocol enforcement plugin for DeepSeek Harness — forces memory_weave before tool calls, auto-ingests session turns, injects memory context into agent steps. Bridges the opencode-memory MCP server with a hard protocol gate.
Send DeepSeek Harness interaction and turn notifications to a Feishu bot
DeepSeek Harness (dsh) plugin: prune the tool schema and inject a structural preamble on a session's first turn only.
An intent-calibrated DSH discussion mode that prevents complex conversations from drifting and turns them into evidence-based next steps.
DeepSeek Harness plugin that turns UI screenshots into implementation-grade web specs using OCR, deterministic geometry, scene graphs, assets, and render comparison.
Fifth DeepSeek Harness mode with configurable Codex, Claude Code, or DSH audits, user-approved remediation turns, progressive Cordis/skill elevation, safety pauses, and Web/TUI review controls.
Long-term memory for DeepSeek Harness agents: automatic recall each turn, a retain/recall/reflect tool, standing rules, and visibility you scope to the whole bank, a preset, or a session.
Cite what matters in DeepSeek Harness: @ a past session, quote earlier messages with comments, and open the files this turn just wrote.
Running-turn TPS badge for DSH Web: live tokens-per-second inside the Deep diving status row
Minimal first-turn conditioning for DeepSeek Harness Web sessions, with a persistent UI toggle.
Two-tier model router for DeepSeek Harness: an LLM Judge sends each turn to the cheap Fast chain or the Smart chain by expected cost, with multi-model fallback chains, exponential-backoff failover, cache-aware switching and task-level orchestration. A DSH adaptation of pi-shift-router.
Replace DeepSeek Harness's 'Deep diving...' turn status with rotating Claude Code-style loading words.
DeepSeek Harness plugin: generate_ui — the model turns a request (+ data) into an A2UI document rendered as an interactive surface in the chat, generated through dsh's own model service (valuz-genui core)
A 'slacking off' mini-games plugin for DeepSeek Harness: five casual games (sliding-number puzzle, Sudoku, snake, Schulte grid, number memory) that pop up when a turn starts, with auto-saved progress across sessions.
Host-level vision bridge for text-only models: analyze_image tool (Ollama local / Xiaomi MiMo cloud / any OpenAI-compatible endpoint) returning structured evidence.
DeepSeek Harness achievement & gamification plugin: cross-session badges for turns, tool calls, sessions and daily streaks, with a badge panel, unlock toasts and a ctx.achievements service. Zero core changes.
对话轮次索引:Web GUI 右侧边栏列出每一轮用户提问,点击跳转到对应位置 (Turn index sidebar for the DeepSeek Harness web GUI: one entry per user turn, click to jump)
DeepSeek Harness 的 TencentDB Agent Memory Adapter:自动 recall、turn 捕获与只读搜索工具。
Turn stock mentions in DSH assistant replies into clickable quote and news buttons.
LLM-as-a-Verifier plugin for DeepSeek Harness: fine-grained logprob-based scoring, pairwise pivot tournaments (best-of-N) and an automatic end-of-turn quality gate.
Model-facing generate_image tool for DeepSeek Harness (dsh): generates brand-new images with the Google Gemini image-generation service via the Antigravity CLI, saves them, and returns the paths. Hot-pluggable — install with `dsh plugin --profile web add dsh-tool-generate-image`.
DSH Web GUI 会话统计插件:在顶部「对话」「轨迹」之后新增「会话统计」tab,列出每个对话的轮数、步数、token 用量(输入/输出/缓存)、模型调用、工具调用与耗时等数据,点击会话可查看逐条调用明细。数据直接解析 ~/.dsh/sessions 下的会话日志(zstd),无需常驻监听,历史对话同样可查。