CacheBoost
Server Details
Warm CDN and reverse-proxy caches: manage sites, run warms, read cache-hit stats.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 9 tools
Most tools map cleanly to a distinct resource+action (sites, runs, config checks, identity). The one soft spot is the run trio: get_run, list_runs, and list_warm_runs all return run status, and the descriptions have to explicitly steer the agent ("prefer list_warm_runs or get_run" for a run you just created, list_runs is "the history view"), which signals real overlap between 'boost runs' and 'warm runs'.
A clear verb_noun pattern dominates (get_site, list_sites, get_run, list_runs, warm_site, get_config_check), so tools are predictable. Minor deviations: check_cache_config uses 'check' as the verb rather than a start/create style, and its object name 'cache_config' doesn't match the paired 'config_check' used in get_config_check; whoami is an acceptable conventional outlier.
Nine tools is well-scoped for a cache warming + diagnostics service: a starting pair for each workflow (start/check config, warm/list/get run), two site lookups, and a bootstrap identity tool. Nothing looks redundant enough to remove.
Core lifecycle is covered: identify scopes, enumerate/inspect sites, start a warm and follow it to completion, start a config check and read its scored results with an action plan. Gaps are minor but real: no way to cancel or abort an in-flight warm/check, no history listing of past config checks (only by check_id or 'most recent'), and list_sites deliberately does not paginate.
Publisher details
- Operator
- CacheBoost
- Operator website
- https://www.cache-boost.com
- Vendor relationship
- Not applicable
- Documentation
- https://www.cache-boost.com/support/mcp/overview
- Privacy policy
- Not provided
- Trust center
- Not applicable
- Restrictions
- Not applicable
Available Tools
9 toolscheck_cache_configAInspect
Start a cache configuration diagnostic ("config check") of one of your validated sites. CacheBoost sends about 40 test requests to the site (home page and sample pages, mobile and logged-in variants, static assets), adds the site's Cloudflare settings and WordPress plugin snapshot when they are connected, and scores how well pages and assets are cached. The check is asynchronous (usually under a minute): this returns {id, status: "pending"}, then poll get_config_check with that check_id until status is "done" or "error". One check at a time per site and at most 10 per site per 24 hours; it does not consume the warm quota. Requires the sites:write scope.
Inputs: site_id (required).
get_config_checkAInspect
Get a cache configuration diagnostic: pass check_id for one check, or site_id for the site's most recent check whatever its status. Returns status (pending, running, done, error; error_code tells why a check failed), score 0-100 (capped at 49 while a critical criterion fails), coverage (share of criteria that could be evaluated), sources (http, cloudflare, wordpress), the detected stack (CDN, cache plugin, host), counts per status, action_plan (up to 3 criteria to fix first, with the score gain of each) and results: one entry per criterion with criterion_id (e.g. html.edge_hit), status (pass, warn, fail, skip = data unavailable, na = not applicable, info), weight (10 critical, 5 major, 2 minor, 0 info) and evidence {code, params}. Requires the sites:read scope.
Inputs: site_id, check_id.
get_runAInspect
Get the detailed status and cache-hit statistics of a single run. Use it to check whether a run created by warm_site has completed, and how many of its URLs were served from cache. When a post-warm check was requested (validation_ratio), validation_status is verified, to_fix, probable (no cache status header: judged on response times only), not_verified or in_progress (check still running: poll again) and validation holds the counts. Requires the runs:read scope.
Inputs: run_id (required).
get_siteAInspect
Get one site by id (domain, validation status, validation token). Requires the sites:read scope. A site outside the key's allowed site_ids is reported as not found.
Inputs: site_id (required).
list_runsAInspect
List boost runs across all of the key's sites (most recent first). Optionally filter by site_id or boost_id, and paginate with limit/offset. This is the history view: to follow a run you just created with warm_site, prefer list_warm_runs or get_run. Requires the runs:read scope.
Inputs: limit, offset, site_id, boost_id.
list_sitesAInspect
List the sites accessible to the API key (most recent first). Requires the sites:read scope. Takes no arguments: this endpoint returns all authorized sites and does not paginate.
Inputs: no named parameters.
list_warm_runsAInspect
List the inline warm runs of a site (most recent first) to poll their status after warm_site. Each run carries validation_status (verified, to_fix, probable, not_verified, in_progress) and the validation block when a post-warm check was requested. probable means the site sends no cache status header, so the check could only compare response times. Requires the boosts:read scope.
Inputs: site_id (required), limit, offset.
warm_siteAInspect
Warm (pre-cache) a list of same-origin URLs for one of your sites. Creates a warm run and returns its run_id. The run is asynchronous and takes seconds to minutes depending on the URL count, so follow it with list_warm_runs or get_run instead of assuming it finished. A site runs one warm at a time. Requires the boosts:write scope.
Inputs: urls (required), site_id (required), region, variations, +1 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Read SpinupWP servers, sites and events; purge caches, restart services, run git deploys.
Build and update websites on Wirechunk.
Web cron for sites without their own scheduler: calls your URLs on schedule and logs every call.
DNS, TLS, HTTP and SEO checks, plus stateful change watches that diff a page over time.
Related MCP Servers
- AlicenseCqualityDmaintenanceInteract with Webflow sites, pages, and collections.223,649 npm142MIT
- AlicenseAqualityCmaintenanceA unified CLI, web dashboard, and Python API for SLURM job management.1515 PyPI23Apache 2.0
- AlicenseNot gradedqualityCmaintenanceProvides SEO audits, crawling, performance analysis, and deployment management for web projects via a stdio MCP server.4 npmMIT

@tengence/geo-mcpofficial
AlicenseNot gradedqualityBmaintenanceEnables AI clients to manage GEO/SEO content pipelines for multiple sites, including content ingestion, quality checks, WordPress publishing, image handling, monitoring, and search submission via 22 MCP tools over stdio or authenticated HTTP.117 npmMIT
Glama MCP Gateway
Add one secure layer between your agents and this server.