dsh-record-replay: A macOS Desktop Workflow Recording DSH plugin for DeepSeek Harness

dsh-record-replay is a macOS desktop workflow recording plugin (DSH plugin) for DeepSeek Harness, maintained by humblebanana under the MIT license. It turns the Open Record/Replay macOS workflow recorder into first-class harness capabilities: it registers the open-record-replay skill and six model-facing orr_* tools, enabling an agent to learn a user-demonstrated desktop workflow—recording real macOS actions, validating evidence, and packaging it for the host agent's native skill creator.

What core tools does the dsh-record-replay DSH plugin provide?

The plugin offers six orr_* tools covering the full pipeline: orr_permissions_check verifies Accessibility/Input Monitoring before recording; orr_record_start and orr_record_stop begin/end capturing the user's demonstration; orr_session_events reads the evidence stream (events.jsonl, capped at a limit); orr_session_validate checks the recording against the official contract; orr_skill_prepare packages the evidence directory. Additionally, it includes a built-in orr_skill_create fallback that generates a SKILL.md skeleton following the Anthropic skills spec and installs to ~/.agents/skills/<name>/.

How do I install dsh-record-replay?

You need macOS (the recorder's native backend is Swift; requires Xcode Command Line Tools), Node.js >= 22.19, a DeepSeek Harness installation, and an open-record-replay checkout (the plugin invokes its bin/orr.js). Build the package and add it to a profile as a bundle:

git clone https://github.com/<you>/dsh-record-replay.git
cd dsh-record-replay
pnpm install
pnpm build
pnpm pack                      # produces dsh-record-replay-0.1.0.tgz
dsh plugin --profile web add ./dsh-record-replay-0.1.0.tgz

dsh plugin add records the package in the profile's package.json dependencies and dsh.profile.bundles, and the harness heals the profiles/node_modules fallback. The shipped cordis.patch.yml mounts a neutral row; point it at your checkout by overlaying the row from your profile's cordis.patch.yml.

What configuration options does dsh-record-replay have?

Configuration keys include cliPath (explicit path to bin/orr.js, overrides repoRoot), repoRoot (path to an open-record-replay checkout; CLI is <repoRoot>/bin/orr.js), runsOut (recordings directory, default runs), and skillInputsOut (skill-input packages directory, default skill-inputs). The CLI runs with the session workspace as its working directory, so recordings and skill packages land where the agent's filesystem tools can read them.

Which DeepSeek Harness versions and platforms does dsh-record-replay support?

The plugin targets macOS and depends on the DeepSeek Harness runtime (Node.js >= 22.19). It integrates with DSH via official dependencies like @deepseek-ai/dsh-skill, @deepseek-ai/dsh-subprocess, and @deepseek-ai/dsh-tools, categorizing it as an Agent core plugin. The repository was created on 2026-08-13 and last updated on 2026-08-13, with 7 stars and no open issues.

View more DSH plugins at dsh-plugin.org