dsh-dcp
fan56
dsh plugin: deterministic context compression backend — zero LLM calls, reproducible compression
Seetraum94Updated 1mo ago
Deepseek Harness 删除会话
Practical score
81 / 100
Setup
Ready to use
License
MIT
Language
JavaScript
Copy the command, or hand the whole task to your harness agent.
git clone https://github.com/Seetraum/harness-session-delete.gitPlease install the DeepSeek Harness plugin "harness-session-delete" for me. - GitHub repository: Seetraum/harness-session-delete (https://github.com/Seetraum/harness-session-delete) - Type: a Cordis plugin (installed into a DSH profile) - It should work out of…
A cleaned excerpt from the project's own documentation.
dsh-session-recycle-bin Session recycle bin & permanent delete for DeepSeek Harness. Move the current session into a recycle bin from the conversation header, restore or permanently delete archived sessions (single or batch), with live-session protection. The interface automatically follows the browser language: Chinese locales use Chinese and all other locales use English. English 中文 --- ✨ Features - Conversation-header archive button — move the current session into the recycle bin from the header (no confirmation; always recoverable). - No duplicate of DSH's own affordance — archiving from the sidebar uses DSH's built-in hover button; this plugin no longer injects a second, differently worded icon (0.4.4 removed the old row trash icon and the document-wide MutationObserver that drove it). - Recycle-bin manager (Settings → 🗑 会话回收站) — archived sessions with multi-select: batch restore, batch permanent delete; each item shows its ID and path. - True permanent delete — physically removes the session log from disk and the projection-cache record;…
Open on GitHubSimilar tools from the same corner of the ecosystem.
fan56
dsh plugin: deterministic context compression backend — zero LLM calls, reproducible compression
Yuuyuko-uu
🐋 DSH 鱼的记忆:给会话一份会淡忘的长期记忆,每条都带时间。名字里的「鱼」来自 DeepSeek 的鲸鱼。
mengwEX
AI-managed Auto Compact for DeepSeek Harness: the agent decides when to compact context and what to retain based on task progress, without waiting for a fixed token threshold.