Plugin directory / Developer / loongport-dsh
loongport-dsh
Unverified SailingLoong
What it does
LoongPort npm setup CLI for DeepSeek Harness (dsh) OpenAI-compatible routes
Unverified — not yet verified
LoongPort npm setup CLI for DeepSeek Harness (dsh) OpenAI-compatible routes Not yet verified — install and test it yourself.
“Unverified” means our automated CI has not yet installed this plugin. Feature descriptions and version compatibility are the author’s claims. This is not a security audit and not an endorsement of third-party code.
README
loongport
loongport is a DeepSeek Harness
(DSH) Cordis bundle for verified LoongPort OpenAI-compatible providers. It also includes an
advanced CLI for custom endpoints.
Install in DeepSeek Harness
Install the public package into the DSH profile you use:
dsh plugin --profile <profile> add loongport
Open Settings → LoongPort in DSH, then:
- Select a verified provider. The default is DeepSeek's official API at
https://api.deepseek.com/v1; BestAPI (https://api.bestapi.store/v1) is the
verified relay option. - If you need an account, use the provider's Register or Sign in link. It opens the
provider's own website in your browser. - Generate an API key with the provider and paste it into the dialog.
- Select
deepseek-v4-flashordeepseek-v4-pro, then save.
The signed LoongPort directory is the authority for provider identity, URLs, available models,
and any published invitation code. VeriDrop data, when shown, is only a display-only health
observation and cannot change provider configuration. DSH credentials store API keys; the
Settings page only retains whether a key is configured.
Version 0.2.0 does not automate browser authorization, collect browser sessions, or access
cookies/local storage. Registration/login and API-key creation remain on the provider's site.
Advanced custom endpoint
For a manually supplied OpenAI-compatible endpoint, run the CLI directly:
LOONGPORT_API_KEY='your-api-key' npx loongport dsh setup \
--base-url https://relay.example.com/v1 \
--model model-id \
--write
The API key is read only from LOONGPORT_API_KEY; never place it in a command argument. You
can repeat --model to register more than one model.
By default, the command performs a dry run. Add --write to update DSH's
configuration. It uses $DSH_HOME when set, otherwise ~/.dsh, and writes:
settings.yaml: one selectedllm-pi-ai.providers.<route>entry.credentials.yaml: the selected credential name and API key (owner-only)
Other DSH settings, provider routes, and credentials are preserved. The
command uses DSH's built-in llm-pi-ai transport; it does not implement its
own streaming, retries, or tool calls.
Options
--base-url <url> OpenAI-compatible HTTP(S) endpoint (required)
--model <id> Model to expose; repeat for multiple models (required)
--route <name> DSH provider route (default: loongport)
--credential-name <name> DSH credential key (default: LOONGPORT_API_KEY)
--write Apply changes; omitted means dry run
The CLI remains an explicit advanced path. It supports manually provisioned
OpenAI-compatible endpoints, but does not automate third-party browser login, relay discovery,
desktop-app bridging, or automatic tier switching.
Maintainers
Agents and maintainers should read the agent guide
before changing the package and follow the release runbook
before publishing.
License
Install
Install the catalog once, then DeepSeek Harness can find and install any plugin from this site automatically:
dsh plugin add dshbase-catalog Then say "install loongport-dsh for me" — your agent finds it in the directory and installs it. Docs: dshbase-catalog · verified packs.
This plugin is GitHub source (not published to npm) — install it straight from the repo:
Web profile:
dsh plugin --profile web add github:SailingLoong/loongport-dsh Headless (CLI) profile:
dsh plugin --profile headless add github:SailingLoong/loongport-dsh Test report
Not yet L3-verified — see failure note below if we already ran it.