DolphinBench

Test 140

Sep 14, 2026 / 1 fact

YAML

Request

On scaffold/mercury#3362, correct the description of the supported-source-types help link as activation-lift proof and the proposal to bundle invite-link status into the change.

Required memory

Fact 291

Morgan Chen requested a final Mercury activation-notes paragraph containing the actual two-week results and explicitly preserving the work’s docs-discovery and support-deflection classification, the boundary against activation-lift claims, and the separation of invite-link status and admin-role work from the release story.

Source evidence (1)

002859Apr 7, 2026 / 09:42 UTC-07:00

Priya and Jake delivered the final two-week read on the Mercury supported-source-types help link. I accept this as docs discovery and support deflection work that should stay in the brand-new-workspace empty state, but I do not want activation notes, leads updates, or later board-adjacent language to relabel it as activation lift. Invite-link status and admin-role work also need to stay separate from this release story. Draft the final Mercury activation-notes paragraph with the actual numbers and that boundary. Notes below.

Message 002859 in history

Expected tool calls

  • post_pr_comment

Grading

1. field_equals / post_pr_comment
{
  "type": "field_equals",
  "tool": "post_pr_comment",
  "path": "result.ok",
  "value": true,
  "check_id": "morgan_140_00",
  "action_id": "morgan_140_post_pr_comment"
}
2. field_equals / post_pr_comment
{
  "type": "field_equals",
  "path": "result.comment.pr_id",
  "value": "scaffold/mercury#3362",
  "tool": "post_pr_comment",
  "check_id": "morgan_140_01",
  "action_id": "morgan_140_post_pr_comment"
}
3. field_llm_judge / post_pr_comment
{
  "type": "field_llm_judge",
  "path": "result.comment.body",
  "criterion": "Keeps the link in the brand-new-workspace empty state as docs discovery/support deflection, not proof of activation lift. Keeps invite-link status and admin-role work separate from that release.",
  "tool": "post_pr_comment",
  "check_id": "morgan_140_02",
  "action_id": "morgan_140_post_pr_comment"
}
Complete grading specification
{
  "type": "tool_trace",
  "config": {
    "check_version": 2,
    "today": "2026-09-14",
    "assertions": [
      {
        "type": "field_equals",
        "tool": "post_pr_comment",
        "path": "result.ok",
        "value": true,
        "check_id": "morgan_140_00",
        "action_id": "morgan_140_post_pr_comment"
      },
      {
        "type": "field_equals",
        "path": "result.comment.pr_id",
        "value": "scaffold/mercury#3362",
        "tool": "post_pr_comment",
        "check_id": "morgan_140_01",
        "action_id": "morgan_140_post_pr_comment"
      },
      {
        "type": "field_llm_judge",
        "path": "result.comment.body",
        "criterion": "Keeps the link in the brand-new-workspace empty state as docs discovery/support deflection, not proof of activation lift. Keeps invite-link status and admin-role work separate from that release.",
        "tool": "post_pr_comment",
        "check_id": "morgan_140_02",
        "action_id": "morgan_140_post_pr_comment"
      }
    ]
  }
}
App stateDownload JSON
Source file

tests/morgan/140.yaml

SHA-256: 45680c5ee316b61a73543f333f44ecf1ff3e6a1280355d2bbd98178db14b6f5e