Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Manages digital garden notes, link structures, and health metrics. Use when curating a knowledge base, pruning stale notes, or tracking content maturity.
.claude/skills/athola-digital-garden-cultivator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 49% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 1431% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 91% | 0% |
Design, manage, and evolve digital gardens as living knowledge bases. Digital gardens are interconnected note collections that grow organically, emphasizing evolution over perfection.
A digital garden approach to knowledge management that:
bashpython -m memory_palace.garden_metrics path/to/garden.json --format brief
Verification: Run python --version to verify Python environment.
json - Full metrics as JSONbrief - One-line summaryprometheus - Prometheus exposition formatstructures - use memory-palace-architect
structures - use memory-palace-architect
| Level | Status | Description | |-------|--------|-------------| | Seedling | New/rough | Early ideas, incomplete thoughts | | Growing | Developing | Being actively refined | | Evergreen | Mature | Stable, well-developed content |
yamlgarden: sections: ["research", "patterns", "experiments"] plots: - name: "My First Note" purpose: "reference" # reference | evergreen | lab maturity: "seedling" # seedling | growing | evergreen inbound_links: [] outbound_links: [] last_tended: "2025-11-24T10:00:00Z"
Verification: Run the command with --help flag to verify availability.
| Action | Frequency | Purpose | |--------|-----------|---------| | Quick prune | Every 2 days | Remove dead links, fix typos | | Stale review | After 7 days inactive | Assess content freshness | | Archive | After 30 days inactive | Move to archive or delete |
modules/linking-patterns.mdmodules/maintenance.mdmodules/maintenance.mdmemory-palace-architect - Host garden within palace structureknowledge-locator - Search garden contentsession-palace-builder - Seed garden from session insightsPre-commit hooks failing Run SKIP=... git commit to bypass temporarily, then fix issues
Merge conflicts Use git merge --abort to reset, then resolve conflicts carefully
Commit rejected Check hook output and fix reported issues before committing again
python -m memory_palace.garden_metrics runs without error on the targetgarden.json and outputs link density, bidirectional coverage, freshness, and maturity ratio
surfaced for the user's review
seedling, growing, or evergreen
or reconnected during the maintenance pass
garden.json is absent or malformed, the error is reportedwith the path that was searched rather than silently completing
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-09 | pass→pass | 9,045 | 3,061 | -66% | 1 | 1 | 0% | 1,373 | 1,668 | +21% | 0 | 0 | — |
case-10 | fail→pass | 6,590 | 2,037 | -69% | 1 | 1 | 0% | 993 | 1,476 | +49% | 0 | 0 | — |
case-01 | fail→fail | 16,939 | 33,744 | +99% | 1 | 1 | 0% | 3,370 | 6,247 | +85% | 0 | 0 | — |
case-02 | fail→pass | 14,671 | 17,352 | +18% | 1 | 1 | 0% | 2,901 | 4,052 | +40% | 0 | 0 | — |
case-03 | fail→pass | 5,052 | 11,942 | +136% | 1 | 1 | 0% | 212 | 3,245 | +1431% | 0 | 0 | — |
case-04 | pass→pass | 15,249 | 4,406 | -71% | 1 | 1 | 0% | 2,332 | 1,885 | -19% | 0 | 0 | — |
case-05 | fail→pass | 11,153 | 2,746 | -75% | 1 | 1 | 0% | 1,773 | 1,614 | -9% | 0 | 0 | — |
case-06 | fail→pass | 5,045 | 2,102 | -58% | 1 | 1 | 0% | 772 | 1,472 | +91% | 0 | 0 | — |
case-07 | fail→pass | 4,784 | 1,706 | -64% | 1 | 1 | 0% | 778 | 1,410 | +81% | 0 | 0 | — |
case-08 | fail→pass | 8,822 | 1,721 | -80% | 1 | 1 | 0% | 1,478 | 1,409 | -5% | 0 | 0 | — |
case-11 | pass→pass | 9,832 | 2,434 | -75% | 1 | 1 | 0% | 1,578 | 1,550 | -2% | 0 | 0 | — |
case-12 | fail→pass | 10,395 | 2,243 | -78% | 1 | 1 | 0% | 1,685 | 1,552 | -8% | 0 | 0 | — |
case-13 | fail→pass | 11,343 | 2,995 | -74% | 1 | 1 | 0% | 1,889 | 1,658 | -12% | 0 | 0 | — |
case-14 | fail→pass | 7,153 | 2,031 | -72% | 1 | 1 | 0% | 1,022 | 1,449 | +42% | 0 | 0 | — |
case-15 | fail→pass | 13,617 | 2,090 | -85% | 1 | 1 | 0% | 1,986 | 1,463 | -26% | 0 | 0 | — |
case-16 | fail→pass | 11,195 | 1,922 | -83% | 1 | 1 | 0% | 1,978 | 1,436 | -27% | 0 | 0 | — |
case-17 | pass→pass | 6,518 | 1,936 | -70% | 1 | 1 | 0% | 1,108 | 1,507 | +36% | 0 | 0 | — |
case-18 | pass→pass | 10,189 | 2,862 | -72% | 1 | 1 | 0% | 1,607 | 1,676 | +4% | 0 | 0 | — |
case-19 | pass→pass | 6,856 | 2,511 | -63% | 1 | 1 | 0% | 1,268 | 1,594 | +26% | 0 | 0 | — |
case-20 | pass→pass | 5,728 | 1,951 | -66% | 1 | 1 | 0% | 933 | 1,431 | +53% | 0 | 0 | — |
case-21 | fail→pass | 7,481 | 1,427 | -81% | 1 | 1 | 0% | 1,181 | 1,375 | +16% | 0 | 0 | — |
case-22 | fail→pass | 8,159 | 1,427 | -83% | 1 | 1 | 0% | 1,190 | 1,328 | +12% | 0 | 0 | — |
case-23 | fail→pass | 10,782 | 1,966 | -82% | 1 | 1 | 0% | 1,669 | 1,467 | -12% | 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 +65 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.