dsh-deepatlas
Run the following command in DeepSeek Harness:
dsh plugin install Oscar-Williams/dsh-deepatlas
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install Oscar-Williams/dsh-deepatlas in a plugin-capable terminal to install; the full source is available at https://github.com/Oscar-Williams/dsh-deepatlas
About this plugin
Running multi-agent workflows inside DeepSeek Harness often hits the same wall: you need a capability your current profile lacks, but you are unsure which plugin in the ecosystem is the right fit, whether it introduces supply-chain risk, or how to roll back cleanly if something goes wrong. DeepAtlas closes that gap with an auditable, end-to-end workflow that covers capability diagnosis, candidate retrieval, static risk auditing, locked-commit installation, and recovery - with every step recorded locally.
Six tools make up the core. deepatlas_advise compares the current profile against the task and stays silent when coverage is adequate, surfacing one to three suggestions only when a clear gap exists. deepatlas_find searches the local index using normalized capability tags and returns matched evidence, quality scores, and overlap hints. deepatlas_audit performs source-level checks for lifecycle scripts, dependency shapes, native bindings, and Node compatibility, all bound to a full 40-character commit SHA. deepatlas_install executes only after explicit user authorization, creates a profile snapshot before applying changes, and enters a recovery path on failure. Indices, audit caches, and installation records stay on disk in your DSH home directory; no external service is required.
DeepAtlas is built for developers and teams already on DeepSeek Harness 0.1.1 RC with Node 22 or 24. If you maintain plugin combinations across multiple profiles, want to inspect risk signals before installing rather than trusting a README, or need an auditable trail for internal security reviews, the local-first architecture and controlled installation gates are designed for exactly that workflow.
Use Cases
- Filling capability gaps in the current DSH profile such as cross-session memory or message integration
- Running static risk audits and Node compatibility checks before installing third-party plugins
- Maintaining plugin combinations across profiles with auditable local installation records
Best For
- Developers on DeepSeek Harness 0.1.1 RC with Node 22 or 24
- Security and compliance teams evaluating supply-chain risk before plugin installation
- SREs managing plugin combinations across multiple DSH profiles
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.