17,311 tokens across 63 tools — heavy (15–30K). Measured 2026-09-16 under methodology v1.0.
An Anthropic request carries 12,823 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) | Atlassian MCP v3.4.7 |
| status | dynamic |
| tokenizer | tiktoken / o200k_base |
| launch command | uvx mcp-atlassian |
| isolation | docker · ghcr.io/astral-sh/uv:python3.12-bookworm-slim · network bridge · linux/amd64 · network enabled for package fetch; clean FS, no host credentials |
| env vars supplied | JIRA_URL, JIRA_USERNAME, JIRA_API_TOKEN |
| canonical SHA-256 | da19c0b307a393749f9451c0a8ca03344949edc55bf97afc9d3fbdcb26573ef6 |
| category | community |
| source | https://github.com/sooperset/mcp-atlassian |
This server’s
tools/listdiffered between two consecutive captures, so the number is the first capture and moves between sweeps. Treat it as a range, not a constant.
| tool | tokens | share | description | input schema | output schema |
|---|---|---|---|---|---|
| jira_update_proforma_form_answers | 800 | 4.6% | 507 | 145 | 28 |
| jira_update_issue | 602 | 3.5% | 18 | 502 | 28 |
| jira_get_issue | 567 | 3.3% | 54 | 428 | 28 |
| jira_search | 555 | 3.2% | 12 | 461 | 28 |
| jira_create_issue | 550 | 3.2% | 15 | 453 | 28 |
| jira_create_customer_request | 464 | 2.7% | 8 | 370 | 28 |
| jira_get_issue_sla | 463 | 2.7% | 161 | 201 | 28 |
| jira_get_board_issues | 437 | 2.5% | 13 | 338 | 28 |
| jira_search_assignable_users | 434 | 2.5% | 160 | 180 | 28 |
| jira_get_field_options | 402 | 2.3% | 76 | 234 | 28 |
| jira_create_issue_link | 391 | 2.3% | 8 | 299 | 28 |
| jira_create_remote_issue_link | 377 | 2.2% | 56 | 231 | 28 |
| jira_add_worklog | 363 | 2.1% | 10 | 268 | 28 |
| jira_transition_issue | 355 | 2.1% | 9 | 264 | 28 |
| jira_add_comment | 340 | 2.0% | 8 | 251 | 28 |
| jira_move_issue | 312 | 1.8% | 87 | 133 | 28 |
| jira_batch_create_issues | 308 | 1.8% | 8 | 214 | 28 |
| jira_update_version | 306 | 1.8% | 48 | 174 | 28 |
| jira_batch_create_versions | 290 | 1.7% | 9 | 197 | 28 |
| jira_get_issue_development_info | 288 | 1.7% | 51 | 144 | 28 |
| jira_get_agile_boards | 280 | 1.6% | 13 | 180 | 28 |
| jira_get_issues_development_info | 274 | 1.6% | 27 | 152 | 28 |
| jira_batch_get_changelogs | 264 | 1.5% | 12 | 160 | 28 |
| jira_get_sprint_issues | 261 | 1.5% | 6 | 168 | 28 |
| jira_get_project_epic_hierarchy | 261 | 1.5% | 72 | 92 | 28 |
| jira_assign_issue | 256 | 1.5% | 46 | 126 | 28 |
| jira_update_sprint | 251 | 1.4% | 4 | 164 | 28 |
| jira_get_issue_dates | 248 | 1.4% | 38 | 122 | 28 |
| jira_edit_comment | 242 | 1.4% | 9 | 152 | 28 |
| jira_get_create_fields | 242 | 1.4% | 59 | 95 | 28 |
33 smaller tools omitted (6,189 tokens combined) — all of them are in the raw capture.
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 | 17,311 | 63 | not recorded | not recorded | — |
| 2026-08-19 | 17,311 | 63 | not recorded | docker | no change |
| 2026-08-26 | 17,311 | 63 | not recorded | docker | no change |
| 2026-09-03 | 17,311 | 63 | not recorded | docker | no change |
| 2026-09-04 | 17,311 | 63 | 3.4.7 | docker | no change |
| 2026-09-16 | 17,311 | 63 | 3.4.7 | 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/mcp-atlassian/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.