Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use this skill when the user is discussing daily research work, TODOs, plans, standups, meetings, milestones, or general project progress that should be reflected in Obsidian daily notes, plan notes, and hub updates.
.claude/skills/brycewang-stanford-obsidian-research-log/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 99% | 75 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 400% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 54% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -8% | 0% |
Use this skill to keep project planning and daily progress synchronized.
This is a supporting skill under obsidian-project-memory.
Prefer a small number of durable updates over creating extra planning folders.
Daily/YYYY-MM-DD.md01-Plan.md00-Hub.md recent progress block.claude/project-memory/<project_id>.md01-Plan.md when they become durable commitments.Daily/: append dated blocks instead of rewriting history.Daily/ first.Knowledge/, Writing/, or 01-Plan.md instead of creating a default Meetings/ folder.Daily logs are for process.
If a note will still matter after several days or weeks, promote the durable part into Knowledge/, Experiments/, Results/, or Writing/.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 3,644 | 5,292 | +45% | 1 | 1 | 0% | 262 | 1,310 | +400% | 0 | 0 | — |
case-02 | fail→fail | 5,954 | 6,089 | +2% | 1 | 1 | 0% | 1,110 | 1,550 | +40% | 0 | 0 | — |
case-03 | pass→pass | 9,431 | 4,684 | -50% | 1 | 1 | 0% | 1,528 | 1,163 | -24% | 0 | 0 | — |
case-04 | fail→pass | 3,902 | 4,243 | +9% | 1 | 1 | 0% | 696 | 1,074 | +54% | 0 | 0 | — |
case-05 | fail→pass | 7,735 | 3,561 | -54% | 1 | 1 | 0% | 1,381 | 1,023 | -26% | 0 | 0 | — |
case-06 | pass→pass | 7,165 | 2,760 | -61% | 1 | 1 | 0% | 1,450 | 864 | -40% | 0 | 0 | — |
case-07 | fail→pass | 10,116 | 5,112 | -49% | 1 | 1 | 0% | 1,783 | 1,350 | -24% | 0 | 0 | — |
case-08 | pass→pass | 6,137 | 3,467 | -44% | 1 | 1 | 0% | 1,133 | 1,048 | -8% | 0 | 0 | — |
case-09 | fail→pass | 6,754 | 3,693 | -45% | 1 | 1 | 0% | 1,180 | 1,090 | -8% | 0 | 0 | — |
case-10 | fail→pass | 8,624 | 1,873 | -78% | 1 | 1 | 0% | 1,693 | 701 | -59% | 0 | 0 | — |
case-11 | fail→pass | 7,187 | 1,113 | -85% | 1 | 1 | 0% | 1,286 | 505 | -61% | 0 | 0 | — |
case-12 | fail→pass | 11,547 | 1,510 | -87% | 1 | 1 | 0% | 2,176 | 602 | -72% | 0 | 0 | — |
case-13 | fail→pass | 10,727 | 5,253 | -51% | 1 | 1 | 0% | 1,967 | 1,338 | -32% | 0 | 0 | — |
case-14 | pass→pass | 5,848 | 4,028 | -31% | 1 | 1 | 0% | 1,149 | 1,030 | -10% | 0 | 0 | — |
case-15 | fail→pass | 9,653 | 5,418 | -44% | 1 | 1 | 0% | 1,799 | 1,330 | -26% | 0 | 0 | — |
case-16 | fail→pass | 6,353 | 1,715 | -73% | 1 | 1 | 0% | 1,238 | 584 | -53% | 0 | 0 | — |
case-17 | fail→pass | 10,960 | 4,926 | -55% | 1 | 1 | 0% | 1,894 | 1,430 | -24% | 0 | 0 | — |
case-18 | fail→pass | 9,472 | 6,299 | -33% | 1 | 1 | 0% | 1,664 | 1,472 | -12% | 0 | 0 | — |
case-19 | pass→pass | 8,763 | 4,386 | -50% | 1 | 1 | 0% | 1,498 | 1,135 | -24% | 0 | 0 | — |
case-20 | pass→pass | 3,475 | 2,983 | -14% | 1 | 1 | 0% | 728 | 864 | +19% | 0 | 0 | — |
case-21 | pass→pass | 3,579 | 2,763 | -23% | 1 | 1 | 0% | 661 | 772 | +17% | 0 | 0 | — |
case-22 | pass→pass | 2,222 | 1,639 | -26% | 1 | 1 | 0% | 389 | 571 | +47% | 0 | 0 | — |
case-23 | pass→pass | 10,745 | 7,711 | -28% | 1 | 1 | 0% | 1,975 | 1,791 | -9% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted, and 22 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +57 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.