dsh-matt-skills-flow
Run the following command in DeepSeek Harness:
dsh plugin install hollis-openlab/dsh-matt-skills-flow
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install hollis-openlab/dsh-matt-skills-flow in your DeepSeek Harness terminal to install the plugin, whose source is available at https://github.com/hollis-openlab/dsh-matt-skills-flow .
About this plugin
When you bring Matt Skills into daily engineering, what is missing is not another single command but a persistent, crash-recoverable workflow that carries state across sessions and restarts. dsh-matt-skills-flow fills exactly that gap: it consolidates the current phase, installed Skill snapshot, Decisions, Specs, ticket dependency graph, isolated Git Lanes, review findings, human acceptance, and evidence export into one Flow. The Web UI and /matt-flow commands share the same Host service, so switching between them never loses context.
The core pipeline follows an explicit gate chain. At Flow creation the plugin snapshots the user installed Matt Skills and freezes their content digests. Decisions are recorded and iterated before a Spec is generated and approved. Once approved, a dependency-aware Ticket Graph is published to repository-local Markdown by default, or to GitHub Issues in dependency order when the Host config selects the GitHub tracker. Isolated Lane branches and worktrees are then cut from a Flow-owned integration branch; bounded Lane Agents execute, route blocked questions back to the root session, and retry with a fresh packet digest. Completed Lanes integrate sequentially, pass a two-axis Standards and Spec review, and require explicit human acceptance before an Acceptance Receipt is written. The plugin never pushes, opens pull requests, merges the default branch, or deletes dirty worktrees.
It is built for teams or individual engineers who already rely on Matt Skills and want to wire those capabilities into a resumable, auditable, human-in-the-loop engineering pipeline. If you need parallel isolated development lanes, dependency-driven ticket management, and clear acceptance gates, this plugin upgrades Matt Skills from a collection of skills to a full workflow.
Use Cases
- Chain Matt Skills into a persistent delivery pipeline that survives sessions and restarts
- Run parallel isolated Lanes, advance tickets in dependency order, and integrate sequentially
- Resume interrupted review and acceptance state after a Harness restart
Best For
- Engineers who already use Matt Skills and want a coherent end-to-end pipeline
- Small-to-mid teams needing parallel isolated lanes and dependency-driven ticketing
- Engineering leads who value explicit acceptance gates and auditable evidence trails
Related Plugins
A method pack that makes AI coding agents plan against your real baseline, prove completion with fresh evidence, and reduce reworks and unsafe changes.
Turns the DeepSeek Harness session into a captain that builds a durable sub-agent team, splits goals into dependency-aware tasks, and coordinates work via direct messages and a live Web UI.
Gives coding agents design judgment, letting Claude Code, Cursor, and 70+ agents generate and iterate high-quality UI, presentations, and graphics right from the terminal.
Run the Pi ecosystem's plugins on DeepSeek Harness, unmodified, via a compatibility layer that implements Pi's public extension ABI on DSH's native services.