dsh-acp-for-bitfun: Integrating BitFun into DSH as a Subagent via ACP
dsh-acp-for-bitfun is a DeepSeek Harness plugin maintained by bobleer that registers BitFun as a subagent provider for dsh via the Agent Client Protocol v1 over stdio. After installation, models can delegate tasks to BitFun for execution using the subagent_bitfun tool. The article explains the working principle, prerequisites (BitFun CLI, dsh >= 0.1.0-rc.6, pnpm), installation commands, configuration items in cordis.patch.yml, and verification steps, while also describing the permission policy and fail loud behavior.
Read MoreIntegrate DeepSeek Harness with ACP clients such as Zed using deepseek-harness-acp
The community plugin deepseek-harness-acp (openma-ai, Apache-2.0) turns DeepSeek Harness into an editor-oriented ACP server, which is different from the official automation package @deepseek-ai/dsh-acp. This article verifies via the directory page, GitHub README and npm: the shared credential $DSH_HOME, directory installation commands, two connection methods of standalone dsh-acp and acp profile, as well as the boundaries of workspace-write / danger-full-access.
Read MoreIntegrate BitFun into DeepSeek Harness Sessions via dsh-acp-for-bitfun
`dsh-acp-for-bitfun` is a DeepSeek Harness development and runtime plugin maintained by bobleer, licensed under MIT. It connects the `bitfun acp` command of BitFun CLI to the official dsh subagent client via ACP v1 over stdio, allowing any session to delegate tasks to BitFun using `subagent_bitfun`. The installation command is `dsh plugin add github:bobleer/dsh-acp-for-bitfun`. It requires BitFun >= 0.2.17 and dsh >= 0.1.0-rc.6. The plugin runs with the permissions of the current dsh process, and you should inspect the source code and license before installation.
Read MoreUse billion-context-dsh to let the model handle the context compression of DeepSeek Harness by itself
### 正式翻译： billion-context-dsh is a DSH session plugin maintained by Tyan66666, which implements Active Context Pruning as a CompactionEngine backend: the model calls `compress` to generate its own summaries, and the automatic strategy only nudges. The original text remains in the session log, and decompression and search are both supported. This article verifies the installation commands, mounting methods and differences from the built-in automatic compression according to the directory page, GitHub and npm v0.2.2.
Read More