LayneChai/superpowers-dsh

Superpowers skills for DeepSeek Harness: TDD, debugging, planning, and collaboration skills adapted from obra/superpowers

View on GitHub ↗Jump to charts ↓Open shareable report →

Data as of . Signed-in members get hourly updates — create a free account.

Summary Information

Updated 2 hours ago
Added to GitGenius on September 30th, 2026
Created on August 14th, 2026
Open Issues & Pull Requests: 0 (+0)
GitHub issues: Enabled
Number of forks: 9
Total Stargazers: 104 (+0)
Total Subscribers: 0 (+0)

Repository Insights (GitGenius)

Median issue/PR response: 9.5 days
Mean response time: 14.3 days
90th percentile: 28.0 days
Tracked items: 5

Charts & Analytics

Fetching additional details & charts...

Issue Activity (beta)

Open issues: 0
New in 7 days: 0
Closed in 7 days: 0
Avg open age: N/A days
Stale 30+ days: 0
Stale 90+ days: 0

Recent activity

Opened in 7 days: 0
Closed in 7 days: 0
Comments in 7 days: 0
Events in 7 days: 0

Top labels

No label distribution available yet.

Most active issues this week

Sign in to see which issues are moving.
Sign in

Detailed Description

Superpowers-dsh is a plugin package for DeepSeek Harness that ports core AI agent skills to the DSH platform.

The project adapts the skill framework from obra/superpowers, which provides Claude-Code capabilities for test-driven development, debugging, planning, and collaboration. It registers a skill provider at the host layer of the ctx.skills registry, making fourteen skills available to every agent preset in a user's profile. Skill definitions are bundled with the package and located via import.meta.url, requiring no user configuration beyond installation. Users load skills within agent conversations using the skill tool, with using-superpowers serving as the entry point.

Installation happens through the dsh plugin command, which integrates the package into a DeepSeek Harness profile. The tool supports multiple installation methods: direct npm installation via dsh plugin add superpowers-dsh, installation from GitHub, tarball, or local folders, and agent-assisted installation where users can ask DeepSeek Harness to handle the setup automatically. The package is published to npm and mirrored on npmmirror for faster access in mainland China. After installation, users must restart their profile for the bundle layer to mount and skills to appear in the agent skill directory. The project provides verification steps to confirm successful registration and supports use across different profile types including headless, tui, and custom configurations.

The development activity shows a focused, single-purpose implementation. The project maintains a straightforward codebase centered on skill registration and bundling within the Cordis plugin architecture. Documentation is provided in both English and Simplified Chinese, reflecting attention to accessibility for different user communities. The installation and verification workflows are designed to be user-friendly, with explicit guidance for common scenarios and troubleshooting steps included in the README.