AI Agent Hub
Back to plugins
dsh-all-search preview

dsh-all-search

Web Tools Updated 2026.08.22

Run the following command in DeepSeek Harness:

dsh plugin install RealAlexandreAI/dsh-all-search

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install RealAlexandreAI/dsh-all-search in DeepSeek Harness to install this plugin; the source repository is available at https://github.com/RealAlexandreAI/dsh-all-search.

About this plugin

DeepSeek Harness ships Exa, Perplexity, and built-in DeepSeek search, but wiring multiple backends such as Exa, Tavily, Firecrawl, and Context7 into one search pipeline traditionally means juggling separate API keys and per-service credentials. dsh-all-search registers the AnySearch gateway into ctx.web so a single API key fans out to every backend at once, eliminating the credential-management overhead.

With api_key set, every query routes through the AnySearch MCP gateway. Without it, the same anysearch provider transparently falls back to Firecrawl /v1/search in keyless mode, so the provider never becomes a silent no-op. When both firecrawl_api_key and an AnySearch key are configured, developer-intent queries about repositories, issues, pull requests, or skills hit the Firecrawl Developer Index first, a semantic artifact index spanning READMEs, issues, PRs, and OpenAPI specs, and fall back to AnySearch only when the index returns nothing. Keys stay in the local config file and are never written to logs.

It suits engineering teams running behind a shared NAT or CI egress IP, developers who would rather manage one search credential than several, and automated pipelines that need developer-documentation retrieval folded into a single Harness web-search step.

Screenshots

Use Cases

  • Fan out to Exa, Tavily, Firecrawl, and Context7 through a single AnySearch gateway key
  • Fall back to keyless Firecrawl search when no AnySearch key is configured, keeping the provider always available
  • Route developer-intent queries through the Firecrawl Developer Index semantic artifact search before falling back to AnySearch

Best For

  • Engineering teams that need unified credential management across multiple search backends
  • Ops and platform teams running automated search pipelines behind shared NAT or CI egress IPs
  • Developers who want developer-documentation retrieval folded into a single Harness web-search step