// MCP TOKEN COST

The ABAP-ADT-API MCP-Server MCP server costs 7,239 tokens before the first call.

Connect ABAP-ADT-API MCP-Server and its 127 tool definitions are loaded into the model's context on every request — 3.6% of a 200k window spent before your agent does anything.

QUICK ANSWER The ABAP-ADT-API MCP-Server MCP server's tool definitions consume 7,239 tokens — 3.8× the median MCP server (1,905 tokens). A scoped grant exposing only the tools you use cuts that roughly in proportion.

MEASURED FROM SCHEMAS 127 tools · 7,239 tokens · 3.6% of 200k · 0.7% 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 3.6%

1M WINDOW 0.7%

Corpus context: ABAP-ADT-API MCP-Server ranks #875 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 7,239 tokens go.

Each row is one tool definition as a tools/list entry — name, description and input schema — counted with o200k_base. Average: 57 tokens per tool.

Tool	Category	Tokens	% of server
debuggerSetBreakpoints	Write	191	2.6%
nodeContents	Read	154	2.1%
debuggerListen	Read	124	1.7%
gitCreateRepo	Write	123	1.7%
tableContents	Read	116	1.6%
debuggerDeleteBreakpoints	Destructive	114	1.6%
ddicElement	Read	111	1.5%
switchRepoBranch	Write	109	1.5%
gitPullRepo	Read	108	1.5%
atcWorklists	Read	106	1.5%
debuggerListeners	Read	102	1.4%
activateByName	Write	99	1.4%
activateObjects	Write	93	1.3%
debuggerAttach	Write	92	1.3%
pushRepo	Write	87	1.2%
transportReference	Read	86	1.2%
createTransport	Write	86	1.2%
createAtcRun	Write	85	1.2%
runQuery	Execute	84	1.2%
createObject	Write	84	1.2%
renameEvaluate	Write	83	1.1%
debuggerDeleteListener	Destructive	80	1.1%
findDefinition	Read	80	1.1%
gitExternalRepoInfo	Read	78	1.1%
transportRelease	Read	78	1.1%
syntaxCheckCode	Read	77	1.1%
deleteObject	Destructive	76	1.0%
transportInfo	Read	75	1.0%
debuggerStep	Read	74	1.0%
remoteRepoInfo	Read	73	1.0%
stageRepo	Read	71	1.0%
searchObject	Read	70	1.0%
createTestInclude	Write	70	1.0%
setTransportsConfig	Write	70	1.0%
checkRepo	Read	68	0.9%
tracesDbAccess	Read	65	0.9%
tracesHitList	Read	65	0.9%
abapDocumentation	Read	63	0.9%
atcChangeContact	Read	61	0.8%
codeCompletionFull	Read	61	0.8%
debuggerSetVariableValue	Write	61	0.8%
unitTestRun	Read	60	0.8%
bindingDetails	Read	59	0.8%
revisions	Read	59	0.8%
tracesStatements	Read	59	0.8%
transportsByConfig	Read	59	0.8%
unitTestEvaluation	Read	59	0.8%
unPublishServiceBinding	Write	59	0.8%
packageSearchHelp	Read	58	0.8%
renamePreview	Write	58	0.8%
userTransports	Read	57	0.8%
lock	Write	57	0.8%
publishServiceBinding	Write	57	0.8%
setObjectSource	Write	57	0.8%
extractMethodEvaluate	Read	56	0.8%
unLock	Write	56	0.8%
objectStructure	Read	55	0.8%
unitTestOccurrenceMarkers	Read	55	0.8%
setPrettyPrinterSetting	Write	55	0.8%
codeCompletionElement	Read	54	0.7%
transportAddUser	Write	54	0.7%
transportSetOwner	Write	54	0.7%
debuggerGoToStack	Read	52	0.7%
codeCompletion	Read	51	0.7%
fixProposals	Read	51	0.7%
atcContactUri	Read	49	0.7%
debuggerStackTrace	Read	49	0.7%
usageReferences	Read	49	0.7%
atcCheckVariant	Read	48	0.7%
debuggerChildVariables	Read	47	0.6%
dumps	Read	47	0.6%
atcExemptProposal	Read	45	0.6%
collectionFeatureDetails	Read	44	0.6%
ddicRepositoryAccess	Read	44	0.6%
isProposalMessage	Read	44	0.6%
mainPrograms	Read	44	0.6%
prettyPrinter	Read	44	0.6%
tracesListRequests	Read	44	0.6%
findCollectionByUrl	Read	43	0.6%
findObjectPath	Read	43	0.6%
fragmentMappings	Read	43	0.6%
getObjectSource	Read	43	0.6%
atcRequestExemption	Read	42	0.6%
extractMethodExecute	Read	42	0.6%
getTransportConfiguration	Read	42	0.6%
gitUnlinkRepo	Read	42	0.6%
tracesList	Read	42	0.6%
tracesDeleteConfiguration	Destructive	41	0.6%
extractMethodPreview	Read	41	0.6%
featureDetails	Read	41	0.6%
debuggerVariables	Read	40	0.6%
renameExecute	Write	40	0.6%
tracesDelete	Destructive	38	0.5%
syntaxCheckCdsUrl	Read	38	0.5%
debuggerSaveSettings	Write	38	0.5%
tracesCreateConfiguration	Write	38	0.5%
fixEdits	Read	37	0.5%
tracesSetParameters	Write	37	0.5%
transportDelete	Destructive	36	0.5%
classComponents	Read	36	0.5%
classIncludes	Read	35	0.5%
objectRegistrationInfo	Read	35	0.5%
validateNewObject	Read	34	0.5%
usageReferenceSnippets	Read	32	0.4%
adtCompatibiliyGraph	Read	31	0.4%
atcUsers	Read	30	0.4%
reentranceTicket	Read	30	0.4%
runClass	Execute	29	0.4%
atcCustomizing	Read	29	0.4%
adtCoreDiscovery	Read	28	0.4%
gitRepos	Read	28	0.4%
prettyPrinterSetting	Read	28	0.4%
systemUsers	Read	28	0.4%
hasTransportConfig	Read	27	0.4%
syntaxCheckTypes	Read	27	0.4%
adtDiscovery	Read	26	0.4%
feeds	Read	26	0.4%
healthcheck	Read	26	0.4%
inactiveObjects	Read	26	0.4%
createTransportsConfig	Write	26	0.4%
dropSession	Destructive	25	0.3%
annotationDefinitions	Read	25	0.3%
objectTypes	Read	25	0.3%
transportConfigurations	Read	25	0.3%
login	Write	25	0.3%
loadTypes	Read	24	0.3%
logout	Read	24	0.3%

// SCOPED GRANT

Most agents use a handful of these tools. They pay for all 127.

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 (57 tokens each).

Grant scope	Definition cost	Reduction
All 127 tools (no gateway)	7,239 tokens	—
3 granted tools	~171 tokens	−98%
5 granted tools	~285 tokens	−96%
10 granted tools	~570 tokens	−92%

Model your own stack in the token-cost calculator, or see the ABAP-ADT-API MCP-Server policy for what a sensible grant looks like.

//FAQ

ABAP-ADT-API MCP-Server token-cost questions.

How many tokens does the ABAP-ADT-API MCP-Server MCP server use?+

Its 127 tool definitions total 7,239 tokens — 3.6% 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 ABAP-ADT-API MCP-Server 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 ABAP-ADT-API MCP-Server's token usage?+

Expose fewer tools. A PolicyLayer grant scopes ABAP-ADT-API MCP-Server 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 171 tokens, a 98% 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 127 catalogued ABAP-ADT-API MCP-Server tools. Counts refresh with every site build.