Loading ToolBench

Reading the current corpus…

firescrapling MCP server · F · ToolBench

← Browse

firescrapling

github.com/grayaa/firescrapling

MCP server providing web scraping tools (scrape, crawl, map, extract_media, fetch_savings) over a REST API backend

2026-07-28+STDIO6 tools

48/100

Definition Quality61

Protocol Readiness50

Supportability25

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence

Current-spec patterns detected

Summary

FireScrapling defines 6 tools with complete input schemas and descriptions. However, descriptions are minimal (10-60 chars, well below the 194-char baseline), lack context on WHEN to use each tool, and omit error recovery guidance. Parameter descriptions are present but terse. Output schemas are not documented. Tools lack error handling patterns and recovery guidance. No tool annotations (readOnlyHint, etc.) despite all being read-only. Naming is clear (verb_noun pattern) but descriptions do not explain dependencies or prerequisites.

Tools (6)

crawlread onlyauthsource verified63/100

Start a BFS crawl job. Returns job id; poll with crawl_status.

crawl_statusread onlyauthsource verified60/100

Get crawl job status and paginated page results.

extract_mediaread onlyauthsource verified62/100

Extract media/catalog manifests (URLs only — no proxy/download) for supported sites.

fetch_savingsread onlyauthsource verified55/100

Estimated fetch-credit savings vs always using the expensive tier.

mapread onlyauthsource verified63/100

Discover links on a site (map).

scraperead onlyauthsource verified63/100

Scrape a URL into LLM-ready markdown (or other formats).

Issues (5)

HIGHtool-description ↗

Descriptions are 10 - 60 characters, far below the 194-char baseline. They lack WHEN to use each tool, prerequisites, and error recovery guidance. E.g., 'Scrape a URL into LLM-ready markdown' does not explain when to call scrape vs crawl, or what happens on network errors.

scrapecrawlcrawl_statusmapextract_mediafetch_savings

HIGHresponse-shaper ↗

Output schemas are not documented. LLMs cannot plan downstream calls or extract required fields. E.g., crawl_status returns paginated results but the response structure (fields, pagination tokens, total count) is not visible in the tool definition.

scrapecrawlcrawl_statusmapextract_mediafetch_savings

MEDIUMtool-annotations ↗

No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite all tools being read-only. Agents cannot infer safety properties without explicit hints.

scrapecrawl

Recommendations

Expand tool descriptions to 100 - 200 characters. Include WHEN to use (e.g., 'Use scrape for single-page extraction; use crawl for multi-page discovery'), prerequisites, and common error scenarios. Example: 'Scrape a single URL into LLM-ready markdown. Use this for one-off page extraction; use crawl() for multi-page discovery. Returns markdown or other formats. Fails on invalid URLs, timeouts, or 403 responses, retry with a different URL or check robots.txt.'
Document output schemas for all tools. Specify field names, types, and pagination structure. E.g., crawl_status returns {job_id, status, pages: [{url, title, content, ...}], total_count, offset, limit}.
Add tool annotations: set readOnlyHint=true for all 6 tools to signal they are safe to call without user confirmation.
Expand parameter descriptions with format, range, and validation rules. E.g., 'limit: Max pages to crawl (1 - 1000, default 20). Larger limits increase latency.' 'url: Full HTTP(S) URL (e.g., https://example.com). Must be publicly accessible.'
Add error recovery guidance to descriptions. E.g., 'If crawl fails with timeout, try reducing limit or max_depth. If it fails with 403, the site may block automated access.'

Score history

Overall score trend

First recorded score · v2 rubric

48/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	48	2026-07-28+	v2