// MCP TOKEN COST

The BorealHost MCP server costs 17,722 tokens before the first call.

Connect BorealHost and its 95 tool definitions are loaded into the model's context on every request — 8.9% of a 200k window spent before your agent does anything.

QUICK ANSWER The BorealHost MCP server's tool definitions consume 17,722 tokens — 9.3× the median MCP server (1,905 tokens). A scoped grant exposing only the tools you use cuts that roughly in proportion.

MEASURED FROM SCHEMAS 95 tools · 17,722 tokens · 8.9% of 200k · 1.8% of 1M Method →

// CONTEXT WINDOW SHARE

What that buys before your agent starts working.

Tool definitions are overhead: they occupy context on every request and compete with your code, documents and conversation history for the same window.

200K WINDOW 8.9%

1M WINDOW 1.8%

Corpus context: BorealHost ranks #64 of 3,213 measured MCP servers by definition cost. The median is 1,905 tokens, p90 is 7,952, and the heaviest (Fusionauth) is 183,337 — 92% of a 200k window on its own.

// PER-TOOL BREAKDOWN

Where the 17,722 tokens go.

Each row is one tool definition as a tools/list entry — name, description and input schema — counted with o200k_base. Average: 187 tokens per tool.

Tool	Category	Tokens	% of server
register_domain	Write	695	3.9%
complete_checkout	Write	505	2.8%
install_app	Write	373	2.1%
manage_dns	Write	361	2.0%
create_alert_rule	Write	358	2.0%
add_domain_dns	Write	348	2.0%
get_checkout_status	Destructive	322	1.8%
update_checkout	Write	312	1.8%
create_api_key	Write	308	1.7%
register	Write	300	1.7%
request_api_key	Destructive	285	1.6%
create_checkout	Write	265	1.5%
get_logs	Read	262	1.5%
domain_settings	Write	252	1.4%
create_b2_snapshot	Write	244	1.4%
add_firewall_rule	Write	243	1.4%
create_snapshot	Write	238	1.3%
run_malware_scan	Execute	235	1.3%
list_files	Read	225	1.3%
list_plans	Read	224	1.3%
update_account	Write	224	1.3%
add_ssh_key	Write	222	1.3%
create_ftp_account	Write	221	1.2%
database_search_replace	Read	217	1.2%
claim_api_key	Write	217	1.2%
upload_file	Write	212	1.2%
toggle_module	Write	201	1.1%
rotate_key	Write	198	1.1%
scale	Execute	195	1.1%
link_domain	Write	195	1.1%
get_site_status	Read	194	1.1%
cloudflare_purge_cache	Destructive	193	1.1%
cloudflare_set_proxy	Write	191	1.1%
read_file	Read	188	1.1%
execute_query	Execute	186	1.0%
schedule_snapshot	Write	185	1.0%
list_domain_dns	Read	183	1.0%
domain_detail	Read	182	1.0%
whoami	Write	180	1.0%
get_app_status	Read	177	1.0%
set_api_key	Write	174	1.0%
write_file	Write	174	1.0%
get_metrics	Read	171	1.0%
get_resource_snapshot	Read	165	0.9%
manage_plugin	Write	165	0.9%
get_ssh_info	Read	164	0.9%
restore_backup	Write	163	0.9%
delete_file	Destructive	162	0.9%
rollback_snapshot	Read	162	0.9%
list_api_keys	Read	158	0.9%
manage_theme	Write	158	0.9%
decommission	Destructive	157	0.9%
delete_domain_dns	Destructive	154	0.9%
list_snapshots	Read	154	0.9%
search_domain	Read	153	0.9%
cloudflare_proxy_status	Read	152	0.9%
get_stack_info	Read	152	0.9%
create_directory	Write	152	0.9%
list_modules	Read	151	0.9%
list_firewall_rules	Read	148	0.8%
add_cron	Write	148	0.8%
list_backups	Read	142	0.8%
wp_check_updates	Read	142	0.8%
list_subscriptions	Read	140	0.8%
get_billing_portal	Read	139	0.8%
get_snapshot_usage	Read	135	0.8%
cancel_scheduled_snapshot	Destructive	134	0.8%
revoke_api_key	Destructive	134	0.8%
list_alert_rules	Read	134	0.8%
list_php_versions	Read	132	0.7%
remove_firewall_rule	Destructive	130	0.7%
switch_php	Write	130	0.7%
delete_snapshot	Destructive	129	0.7%
ssl_info	Read	128	0.7%
create_backup	Write	125	0.7%
delete_cron	Destructive	123	0.7%
list_apps	Read	123	0.7%
list_tables	Read	123	0.7%
list_themes	Read	123	0.7%
get_database_info	Read	122	0.7%
list_cron	Read	121	0.7%
list_plugins	Read	117	0.7%
delete_alert_rule	Destructive	115	0.6%
remove_ftp_account	Destructive	114	0.6%
cache_status	Read	113	0.6%
wp_update_all	Write	113	0.6%
delete_account	Destructive	107	0.6%
list_ftp_accounts	Read	107	0.6%
deploy	Execute	106	0.6%
cache_toggle	Write	106	0.6%
list_domains	Read	97	0.5%
optimize_database	Write	94	0.5%
ssl_renew	Write	94	0.5%
list_databases	Read	93	0.5%
cache_flush	Write	84	0.5%

// SCOPED GRANT

Most agents use a handful of these tools. They pay for all 95.

A PolicyLayer grant exposes only the tools you allow — ungranted definitions are filtered out of the tool list, so they never enter the context window. Estimates below assume typical-weight tools (187 tokens each).

Grant scope	Definition cost	Reduction
All 95 tools (no gateway)	17,722 tokens	—
3 granted tools	~560 tokens	−97%
5 granted tools	~933 tokens	−95%
10 granted tools	~1,865 tokens	−89%

Model your own stack in the token-cost calculator, or see the BorealHost policy for what a sensible grant looks like.

//FAQ

BorealHost token-cost questions.

How many tokens does the BorealHost MCP server use?+

Its 95 tool definitions total 17,722 tokens — 8.9% of a 200k context window — measured with tiktoken o200k_base over the serialised tools/list payload. Exact counts vary slightly by client and model.

Why does BorealHost consume tokens before I send a message?+

MCP clients load every connected server's tool definitions — name, description, and input schema — into the model's context so it knows what it can call. That payload is charged against your context window on every request, whether or not a tool is used.

How do I reduce BorealHost's token usage?+

Expose fewer tools. A PolicyLayer grant scopes BorealHost to only the tools you allow — ungranted definitions are filtered out of the tool list, so they never enter the context window. A grant of 3 typical tools costs roughly 560 tokens, a 97% reduction.

Does deferred tool loading fix this?+

Partially, in some clients. Claude Code defers MCP tool schemas behind a tool-search step by default, and VS Code has experimental grouping — but you still pay tokens per search and reload, and Cursor, Windsurf and Gemini CLI load definitions upfront. Reducing the exposed tool set cuts the cost in every client.

// METHOD & REVIEW

How these numbers were measured.

01

Serialisation

Each tool is serialised as a tools/list entry — name, description, input schema — from the schemas in the PolicyLayer scan database. Clients differ slightly in framing, so treat counts as close estimates.

02

Tokeniser

tiktoken o200k_base (GPT-4o/o-series). Anthropic's current tokeniser isn't published, so Claude's exact counts will differ; for English text and JSON schemas the totals are close enough to treat these as estimates.

03

Deferred loading

Some clients now defer schema loading (Claude Code's tool search; VS Code experimental grouping). You still pay per search and reload — and Cursor, Windsurf and Gemini CLI load everything upfront.

Computed 07-06-2026 from the PolicyLayer scan database over all 95 catalogued BorealHost tools. Counts refresh with every site build.