liustack/modsearch
Web search bridge for text-only agents: ask the web or X, get structured JSON evidence (search, fetch, citations).
Install
via npm
Resolves a published tarball rather than cloning the repository, and installs without any extra setup. Swap `web` for your profile name if you run another one.
via GitHub · npm package
Installing from GitHub runs the project's build script, which pnpm blocks until you allowlist it — run the command once and pnpm prints the exact key to add under `allowBuilds` in ~/.dsh/profiles/web/pnpm-workspace.yaml.
What happened when we ran it
Installed cleanly when we ran it
Every command here is run in a throwaway container against a clean profile, and the result is whatever the harness recorded — not a guess from the source. Last run 13d ago.
Show it in your README
For maintainers: the badge serves this listing's latest sandbox verdict, so a re-run updates it on its own — and it links readers to the full result here.
Due diligence
What it is — ModSearch is a web search bridge for text-only agents, enabling web search, X search, and returning structured JSON evidence.
Who it is for — It suits users driving text agents in DeepSeek Harness for information retrieval tasks. Users with native web access capabilities can choose not to install it.
Watch out — Sandbox testing passed: installed in a new profile and registered by harness. The plugin will execute shell commands. No obvious pitfalls found.
The verdict — I would install it because it supports automatic failover.
Generated by grok-4.6, and a starting point rather than a verdict. Where it says a plugin installs or does not, that is from a real run in a clean profile — everything else is read off the repository. Trust the source over this.
modsearch is a DeepSeek Harness plugin that adds web, X, and single-page reading tools for text-only agents. It connects the harness’s built-in web_search to an engine chain, adds x_search and read_page, and returns structured JSON evidence with sources and citations. The available engines are Antigravity CLI, Tavily, Exa, Firecrawl, Grok Build, and the built-in local fetcher; configured engines fail over when one fails or exhausts its quota.
Configuration is stored in ~/.modsearch/config.json with mode 0600. API credentials can be set with modsearch config set tavily.apiKey, modsearch config set exa.apiKey, or modsearch config set firecrawl.apiKey, or supplied through TAVILY_API_KEY, EXA_API_KEY, and FIRECRAWL_API_KEY. The plugin can also use compatible endpoints through each engine’s baseURL setting.
It is intended for agents that need current web evidence, X search, or focused page extraction beyond the harness’s native search. It is a poor fit when existing web_search is sufficient, or when external engine setup is unacceptable: Antigravity CLI needs browser sign-in, X search needs Grok Build with SuperGrok or X Premium, and API engines require quotas or credentials. The terminal surface reaches the host terminal to install or run helper CLIs such as agy and grok, and to run the plugin’s configuration commands.
modsearch uses an engine chain. The default channel is Antigravity CLI, which is free but requires browser sign-in. API credentials may be stored in ~/.modsearch/config.json with file mode 0600, or supplied through environment variables.
| Engine | Capability | Configuration |
|---|---|---|
Antigravity CLI |
Web search and page fetch | Install agy and sign in |
Tavily |
Web search | tavily.apiKey; TAVILY_API_KEY |
Exa |
Web search | exa.apiKey; EXA_API_KEY |
Firecrawl |
Web search and page fetch | firecrawl.apiKey; FIRECRAWL_API_KEY |
Grok Build |
X search | Install grok and sign in |
local |
Page fetch | Built in |
Each engine can be pointed at a compatible third-party or self-hosted service with its baseURL setting. Multiple configured engines are tried in order, with automatic failover.
The documented configuration form is modsearch config set <key> <value>, for example modsearch config set tavily.apiKey <key> or modsearch config set tavily.baseURL <url>. The CLI also provides configuration and health-check functions through its config and doctor commands.
The plugin connects the harness’s built-in web_search to the configured engine chain and supplies x_search for X and read_page for focused single-page reading. Results are returned as structured JSON evidence, including sources and citations. A search or fetch can be triggered through conversation when the agent needs current information or a URL read.
At least one search-capable engine is needed for web search. Antigravity CLI requires the agy installation and browser sign-in; Grok Build requires a SuperGrok or X Premium account. Tavily, Exa, and Firecrawl use their respective quotas or API-compatible endpoints. local provides page fetching only, so it does not replace a web-search engine.
Written from the project's own documentation and kept in sync with it. Where the two disagree, the source is authoritative — read the README on GitHub
Same category
Tencent
BrowserSkill bridge for controlling visible Chrome and Edge Agent Windows from DeepSeek Harness, with native browser tools, accessibility and VOM observations, screenshots, owned multi-session control, and a live Web UI overlay.
No one-line install — this plugin lives inside a larger repository and publishes no npm package.
omdsh-dev
Chrome sidebar extension that lets DSH operate your browser directly, no vision capabilities required.
No one-line install — this plugin lives inside a larger repository and publishes no npm package.
Lum1104
Chrome sidebar extension that lets DSH operate your browser directly, no vision capabilities required.
No one-line install — this plugin lives inside a larger repository and publishes no npm package.
DDDMUC
Free, keyless web search for DSH: 7 engines (DuckDuckGo/Bing/SearXNG free + Exa/Perplexity/DeepSeek paid), auto-failover, settings-page UI with API key inputs and official links, web_fetch, and an engine test tool.
Installed cleanly when we ran it
Tabbit-Browser
Gives DeepSeek Harness control of the Tabbit Browser: auto-loads the tabbit-browser skill on install, detects official Tabbit and Tabbit Browser releases (>= 1.9.0), checks the tabbit-cli persistent runtime, diagnoses the per-platform DSH sandbox mode needed to call the CLI, and downloads the region-matched official installer via a background job when no qualifying version is present.
Installed cleanly when we ran it
Tabbit-Browser
Gives DeepSeek Harness control of the Tabbit Browser: auto-loads the tabbit-browser skill on install, detects official Tabbit and Tabbit Browser releases (>= 1.9.0), checks the tabbit-cli persistent runtime, diagnoses the per-platform DSH sandbox mode needed to call the CLI, and downloads the region-matched official installer via a background job when no qualifying version is present.
Installed cleanly when we ran it
