Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create, move, reorder, or link learning units and documents in the Knowledge base. Use when the user wants to structure content ('create unit Module 3', 'move these docs to lesson 2', 'reorder lessons by logical sequence', 'link doc X to unit Y').
.claude/skills/evolution-foundation-knowledge-organize/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 151% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 104% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 232% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 5% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 167% | 0% |
Group: Curation. Units: create, move, reorder, link documents.
| Name | Type | Required | Description | |---|---|---|---| | action | str | yes | create \| move \| reorder \| link \| update | | connection | str | no | Defaults to first ready | | (action-specific args below) | | | |
createArgs: space_id (or space slug), slug, title, sequence_idx (optional), description, prerequisites (list uuids), metadata (dict).
pythonfrom dashboard.backend.sdk_client import evo unit = evo.post( "/api/knowledge/v1/units", {"space_id": space_id, "slug": slug, "title": title, ...}, headers={"X-Knowledge-Connection": connection}, )
move (document → unit)Args: document_id, unit_id (null to unlink).
pythonevo.patch( f"/api/knowledge/v1/documents/{document_id}", {"unit_id": unit_id}, headers={"X-Knowledge-Connection": connection}, )
reorder (units within space)Args: space_id, ordered_ids (list of unit uuids in desired order).
pythonevo.post( "/api/knowledge/v1/units/reorder", {"space_id": space_id, "ordered_ids": ordered_ids}, headers={"X-Knowledge-Connection": connection}, )
link (N documents → 1 unit)Args: unit_id, document_ids (list).
pythonfor doc_id in document_ids: evo.patch( f"/api/knowledge/v1/documents/{doc_id}", {"unit_id": unit_id}, headers={"X-Knowledge-Connection": connection}, )
update (unit metadata)Args: unit_id, title, description, sequence_idx, prerequisites, metadata.
pythonevo.patch( f"/api/knowledge/v1/units/{unit_id}", {"title": title, "description": description, ...}, headers={"X-Knowledge-Connection": connection}, )
Confirmation per action:
✓ Unit created: {title}
slug: {slug}
space: {space}
sequence_idx: {N}
id: {uuid}For reorder:
✓ Reordered {N} units in space {space}:
1. {title_1}
2. {title_2}
...action → list valid actionsunit_id not found → "Unit not found. Use knowledge-browse scope=units."sequence_idx → suggest reorderwrite scope"| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 7,275 | 4,804 | -34% | 1 | 1 | 0% | 485 | 1,580 | +226% | 0 | 0 | — |
case-02 | fail→pass | 13,186 | 4,547 | -66% | 1 | 1 | 0% | 651 | 1,637 | +151% | 0 | 0 | — |
case-03 | fail→pass | 4,425 | 4,247 | -4% | 1 | 1 | 0% | 766 | 1,561 | +104% | 0 | 0 | — |
case-04 | fail→pass | 4,580 | 8,627 | +88% | 1 | 1 | 0% | 700 | 2,327 | +232% | 0 | 0 | — |
case-05 | fail→pass | 9,818 | 7,232 | -26% | 1 | 1 | 0% | 1,533 | 1,604 | +5% | 0 | 0 | — |
case-06 | fail→pass | 4,379 | 4,987 | +14% | 1 | 1 | 0% | 601 | 1,606 | +167% | 0 | 0 | — |
case-07 | fail→pass | 5,005 | 3,339 | -33% | 1 | 1 | 0% | 913 | 1,406 | +54% | 0 | 0 | — |
case-08 | fail→pass | 6,616 | 2,648 | -60% | 1 | 1 | 0% | 385 | 1,229 | +219% | 0 | 0 | — |
case-09 | fail→pass | 3,246 | 3,670 | +13% | 1 | 1 | 0% | 501 | 1,480 | +195% | 0 | 0 | — |
case-10 | fail→pass | 9,063 | 5,732 | -37% | 1 | 1 | 0% | 1,679 | 1,884 | +12% | 0 | 0 | — |
case-11 | fail→pass | 7,880 | 2,725 | -65% | 1 | 1 | 0% | 1,466 | 1,268 | -14% | 0 | 0 | — |
case-12 | fail→pass | 4,581 | 4,327 | -6% | 1 | 1 | 0% | 835 | 1,363 | +63% | 0 | 0 | — |
case-13 | fail→pass | 7,886 | 2,134 | -73% | 1 | 1 | 0% | 1,371 | 1,089 | -21% | 0 | 0 | — |
case-14 | pass→pass | 12,507 | 2,057 | -84% | 1 | 1 | 0% | 2,317 | 1,104 | -52% | 0 | 0 | — |
case-15 | pass→pass | 8,961 | 3,222 | -64% | 1 | 1 | 0% | 1,529 | 1,317 | -14% | 0 | 0 | — |
case-16 | pass→pass | 5,931 | 1,574 | -73% | 1 | 1 | 0% | 912 | 981 | +8% | 0 | 0 | — |
case-17 | fail→pass | 4,897 | 2,510 | -49% | 1 | 1 | 0% | 833 | 1,237 | +48% | 0 | 0 | — |
case-18 | fail→pass | 6,239 | 3,800 | -39% | 1 | 1 | 0% | 518 | 1,604 | +210% | 0 | 0 | — |
case-19 | fail→pass | 8,309 | 3,965 | -52% | 1 | 1 | 0% | 1,701 | 1,601 | -6% | 0 | 0 | — |
case-20 | fail→pass | 8,862 | 3,033 | -66% | 1 | 1 | 0% | 778 | 1,294 | +66% | 0 | 0 | — |
case-21 | fail→pass | 5,121 | 3,071 | -40% | 1 | 1 | 0% | 851 | 1,351 | +59% | 0 | 0 | — |
case-22 | fail→fail | 5,596 | 4,613 | -18% | 1 | 1 | 0% | 1,070 | 1,621 | +51% | 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 18 counted toward the lift figure. The other 4 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 +77 percentage points is the difference between those two pass rates over the 18 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.