336 tokens across 9 tools — lean (< 1K). Measured 2026-09-16 under methodology v1.0.
An Anthropic request carries 336 of those tokens as tool definitions. What Claude makes of them is not published for this server: its Claude count is missing, or was taken against a capture this measurement has since replaced.
| server (self-reported) | gitlab-mcp-server v0.5.1 |
| status | measured |
| tokenizer | tiktoken / o200k_base |
| launch command | npx -y @modelcontextprotocol/server-gitlab |
| isolation | docker · public.ecr.aws/docker/library/node:22-slim · network bridge · linux/amd64 · network enabled for package fetch; clean FS, no host credentials |
| env vars supplied | GITLAB_PERSONAL_ACCESS_TOKEN |
| canonical SHA-256 | d51fb022687ee403f9710c9fae31ac91dbc796f5c52a3fcdbe0a91c387aa9365 |
| category | official-reference |
| source | https://github.com/modelcontextprotocol/servers-archived |
| tool | tokens | share | description | input schema |
|---|---|---|---|---|
| get_file_contents | 41 | 12.2% | 13 | 16 |
| create_or_update_file | 40 | 11.9% | 11 | 16 |
| push_files | 39 | 11.6% | 12 | 16 |
| create_merge_request | 38 | 11.3% | 10 | 16 |
| fork_repository | 38 | 11.3% | 11 | 16 |
| create_issue | 36 | 10.7% | 9 | 16 |
| create_branch | 36 | 10.7% | 9 | 16 |
| search_repositories | 33 | 9.8% | 5 | 16 |
| create_repository | 33 | 9.8% | 6 | 16 |
Each tool is tokenized on its own, so the parts do not sum exactly to the whole: the array adds its own brackets and commas, and the tokenizer merges tokens across object boundaries. The badge number is always the count of the whole array, never a sum of parts.
| date | tokens | tools | release | measured in | change |
|---|---|---|---|---|---|
| 2026-08-16 | 336 | 9 | not recorded | not recorded | — |
| 2026-08-18 | 336 | 9 | not recorded | docker | no change |
| 2026-08-26 | 336 | 9 | not recorded | docker | no change |
| 2026-09-03 | 336 | 9 | not recorded | docker | no change |
| 2026-09-04 | 336 | 9 | 0.5.1 | docker | no change |
| 2026-09-16 | 336 | 9 | 0.5.1 | docker | no change |
Some of these sweeps predate the
isolationcolumn, so the conditions they were measured under are not on record.
Full series: results/history.csv.
npx -y mcp-context-cost verify results/gitlab/measurement.json
That re-tokenizes the published capture and checks the count and the hash. If it disagrees with the badge, the badge is wrong — open an issue and it gets corrected.