Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →This skill should be used when the user asks to start a new research project, import an existing code-plus-Markdown repository into Obsidian, or bind the current repository to a compact research knowledge base for future syncing.
.claude/skills/brycewang-stanford-obsidian-project-bootstrap/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -27% | 0% |
Bootstrap a project knowledge base for the current repository.
This is a supporting skill.
Use obsidian-project-memory as the main workflow authority. Use this skill only when a repository still needs its initial binding or rebuild.
obsidian-project-memory detects a research-project candidate but no existing binding.Resolve the vault path from one of:
OBSIDIAN_VAULT_PATH.bash python3 "${CLAUDE_PLUGIN_ROOT}/skills/obsidian-project-memory/scripts/project_kb.py" detect --cwd "$PWD"
bash python3 "${CLAUDE_PLUGIN_ROOT}/skills/obsidian-project-memory/scripts/project_kb.py" bootstrap --cwd "$PWD" --vault-path "$OBSIDIAN_VAULT_PATH"
.claude/project-memory/registry.yaml.claude/project-memory/<project_id>.mdResearch/{project-slug}/00-Hub.mdResearch/{project-slug}/01-Plan.mdResearch/{project-slug}/Knowledge/Source-Inventory.mdResearch/{project-slug}/Knowledge/Codebase-Overview.mdKnowledge/, Papers/, Experiments/, Results/, or Writing/,.git, .venv, node_modules, caches, checkpoints, binaries, and other heavy artifacts.00-Hub.md, 01-Plan.md, Knowledge/, Papers/, Experiments/, Results/, Writing/, Daily/, Archive/.python3 is unavailable in the current shell, use the system Python interpreter that can run project_kb.py and say so explicitly.references/BOOTSTRAP-RUNBOOK.md - preflight decisions, failure modes, and post-bootstrap verification| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 12,380 | 4,909 | -60% | 1 | 1 | 0% | 2,333 | 907 | -61% | 0 | 0 | — |
case-02 | fail→fail | 20,971 | 4,859 | -77% | 1 | 1 | 0% | 3,945 | 1,045 | -74% | 0 | 0 | — |
case-03 | fail→fail | 2,618 | 4,342 | +66% | 1 | 1 | 0% | 321 | 958 | +198% | 0 | 0 | — |
case-04 | fail→fail | 9,436 | 6,299 | -33% | 1 | 1 | 0% | 1,963 | 1,212 | -38% | 0 | 0 | — |
case-05 | fail→fail | 3,577 | 2,107 | -41% | 1 | 1 | 0% | 658 | 910 | +38% | 0 | 0 | — |
case-06 | fail→pass | 7,646 | 3,998 | -48% | 1 | 1 | 0% | 1,652 | 1,584 | -4% | 0 | 0 | — |
case-07 | pass→pass | 11,659 | 5,071 | -57% | 1 | 1 | 0% | 2,088 | 1,586 | -24% | 0 | 0 | — |
case-08 | fail→pass | 8,957 | 5,378 | -40% | 1 | 1 | 0% | 1,697 | 1,809 | +7% | 0 | 0 | — |
case-09 | fail→fail | 7,296 | 7,261 | -0% | 1 | 1 | 0% | 1,240 | 2,087 | +68% | 0 | 0 | — |
case-10 | fail→pass | 14,338 | 3,776 | -74% | 1 | 1 | 0% | 2,538 | 1,402 | -45% | 0 | 0 | — |
case-11 | pass→pass | 8,741 | 4,642 | -47% | 1 | 1 | 0% | 1,541 | 1,531 | -1% | 0 | 0 | — |
case-12 | fail→pass | 9,640 | 3,907 | -59% | 1 | 1 | 0% | 1,762 | 1,441 | -18% | 0 | 0 | — |
case-13 | fail→pass | 6,836 | 1,564 | -77% | 1 | 1 | 0% | 1,346 | 981 | -27% | 0 | 0 | — |
case-14 | fail→fail | 4,519 | 2,176 | -52% | 1 | 1 | 0% | 720 | 1,099 | +53% | 0 | 0 | — |
case-15 | fail→pass | 4,135 | 1,987 | -52% | 1 | 1 | 0% | 724 | 1,077 | +49% | 0 | 0 | — |
case-16 | pass→pass | 11,316 | 3,612 | -68% | 1 | 1 | 0% | 2,031 | 1,411 | -31% | 0 | 0 | — |
case-17 | fail→pass | 6,288 | 1,467 | -77% | 1 | 1 | 0% | 1,149 | 933 | -19% | 0 | 0 | — |
case-18 | pass→pass | 5,769 | 2,707 | -53% | 1 | 1 | 0% | 1,006 | 1,145 | +14% | 0 | 0 | — |
case-19 | fail→pass | 10,262 | 4,969 | -52% | 1 | 1 | 0% | 1,984 | 1,660 | -16% | 0 | 0 | — |
case-20 | pass→pass | 4,816 | 3,411 | -29% | 1 | 1 | 0% | 858 | 1,332 | +55% | 0 | 0 | — |
case-21 | pass→pass | 8,748 | 5,934 | -32% | 1 | 1 | 0% | 1,738 | 1,719 | -1% | 0 | 0 | — |
case-22 | pass→fail | 6,551 | 5,998 | -8% | 1 | 1 | 0% | 1,236 | 1,021 | -17% | 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. 22 cases were attempted, and 17 counted toward the lift figure. The other 5 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 +32 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
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.