From 10f8b0a006391ce56e296f6974216056bca79fb8 Mon Sep 17 00:00:00 2001 From: Noaa Engervall Date: Wed, 30 Sep 2026 16:24:32 -0400 Subject: [PATCH] skills/firecrawl: use Firecrawl MCP tools when they are connected The umbrella skill only knew the CLI, so a session with the Firecrawl MCP connector (Claude plugin users in chat and Cowork, where there is no shell) was told to install and run a CLI it cannot reach. The skill now routes search, scrape, developer, paper and provider tasks to the MCP tools when they are present and falls back to the CLI otherwise, so CLI and Cursor users see no change. Co-Authored-By: Claude --- skills/firecrawl/SKILL.md | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) diff --git a/skills/firecrawl/SKILL.md b/skills/firecrawl/SKILL.md index 67c83d0e4b..cadeebe62d 100644 --- a/skills/firecrawl/SKILL.md +++ b/skills/firecrawl/SKILL.md @@ -1,7 +1,7 @@ --- name: firecrawl description: | - Any live-web task via the Firecrawl CLI — including ordinary web research: searching the web, reading or extracting pages, gathering sources, discovering site URLs, bulk extraction, downloading a site, change alerts, or pages needing clicks/login — web only; local files route to firecrawl-parse. For papers use firecrawl-research-index; for library, API, error, or bug questions use firecrawl-developer-index. + Any live-web task via Firecrawl (its MCP tools when connected, otherwise the CLI) — including ordinary web research: searching the web, reading or extracting pages, gathering sources, discovering site URLs, bulk extraction, downloading a site, change alerts, or pages needing clicks/login — web only; local files route to firecrawl-parse. For papers use firecrawl-research-index; for library, API, error, or bug questions use firecrawl-developer-index. allowed-tools: - Bash(firecrawl *) - Bash(npx firecrawl-cli *) @@ -13,6 +13,20 @@ Search, scrape, and interact with the web. Returns clean markdown optimized for Run `firecrawl --help` or `firecrawl --help` for full option details. For app integration or outcome workflows (research briefs, SEO audits, etc.), route to the `firecrawl-build` / `firecrawl-workflows` skills — see [When to Load References](#when-to-load-references). +## Firecrawl MCP tools + +If Firecrawl MCP tools (names starting `firecrawl_`) are available in this session, use them instead of the CLI for what they cover. Don't install, authenticate, or run the CLI for those tasks. + +| Need | MCP tool | +| ------------------------- | ------------------------------------------------------------------------- | +| Find pages on a topic | `firecrawl_search` | +| Get a page's content | `firecrawl_scrape` with a `url` | +| Answer a coding question | `firecrawl_developer_search` | +| Find or read papers | `firecrawl_research_search_papers`, then `firecrawl_research_read_paper` | +| Structured data providers | `firecrawl_find_tools`, then `firecrawl_scrape` with an `alexandria` body | + +The connected server's tool list is the authority, and some connections expose only the tools above. For a need outside that list (map, crawl, interact, monitor, download, parse), use the CLI when a shell is available; otherwise say this connection doesn't support it. Everything below applies when no Firecrawl MCP tools are connected. + ## Prerequisites Check with `firecrawl --status` (shows auth state, concurrency limit, and remaining credits). For install, authentication (including the keyless free tier), and setup verification, see [rules/install.md](rules/install.md). For output handling guidelines, see [rules/security.md](rules/security.md).