Install any skill in seconds. Free to start, no credit card required.
Get Started Free →SOP wrapping vault_add_edge — create a typed relationship between two vault pages.
.claude/skills/yogsoth-ai-wiki-add-edge/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | 349% | 0% |
| case-04 | ✓→✗ | ▼ Worse | 3% | 0% |
| case-22 | ✓→✗ | ▼ Worse | -22% | 0% |
| case-23 | ✓→✗ | ▼ Worse | -8% | 0% |
| case-01 | ✗→✗ | = Same ✗ | 56% | 0% |
Create a typed edge between two pages in the knowledge graph.
<HARD-GATE> Both source and target paths must reference existing files in the vault. Edge type must be one of the 10 defined types in _schema.md. Do NOT create edges to non-existent pages. </HARD-GATE>
vault_add_edge
| Param | Required | Description | |-------|----------|-------------| | source | yes | Source page path (e.g., "concepts/transformer.md") | | target | yes | Target page path (e.g., "concepts/attention.md") | | edge_type | yes | One of: component_of, instance_of, supported_by, contradicts, supersedes, derived_from, addresses, raises, failed_for, related_to | | weight | no | Edge weight 0.0-1.0 (default 1.0) |
vault_add_edge[[dir/slug]] pointing to the target (dir/slug = target path minus .md). Place at a semantically relevant location. If already present, skip. If no suitable location, append a sentence containing the wikilink at end of body.| Relationship | Edge Type | |-------------|-----------| | A is part of B | component_of | | A is an example of B | instance_of | | A provides evidence for B | supported_by | | A conflicts with B | contradicts | | A replaces B | supersedes | | A was built from B | derived_from | | A answers/solves B | addresses | | A creates/implies B (question) | raises | | A didn't work for B | failed_for | | A is related to B (weak) | related_to |
Returns: { success: boolean, edge_type: string, edge_count: number }
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 18,545 | 32,333 | +74% | 1 | 1 | 0% | 2,334 | 3,633 | +56% | 0 | 0 | — |
case-02 | fail→fail | 16,516 | 20,394 | +23% | 1 | 1 | 0% | 352 | 1,272 | +261% | 0 | 0 | — |
case-03 | fail→fail | 10,425 | 16,020 | +54% | 1 | 1 | 0% | 824 | 709 | -14% | 0 | 0 | — |
case-04 | pass→fail | 11,168 | 16,783 | +50% | 1 | 1 | 0% | 1,030 | 1,063 | +3% | 0 | 0 | — |
case-05 | fail→fail | 16,849 | 19,436 | +15% | 1 | 1 | 0% | 434 | 1,515 | +249% | 0 | 0 | — |
case-06 | fail→fail | 15,892 | 17,636 | +11% | 1 | 1 | 0% | 283 | 1,063 | +276% | 0 | 0 | — |
case-07 | fail→fail | 8,962 | 16,598 | +85% | 1 | 1 | 0% | 783 | 982 | +25% | 0 | 0 | — |
case-08 | fail→fail | 25,090 | 19,267 | -23% | 1 | 1 | 0% | 424 | 996 | +135% | 0 | 0 | — |
case-09 | fail→pass | 15,379 | 18,122 | +18% | 1 | 1 | 0% | 274 | 1,231 | +349% | 0 | 0 | — |
case-10 | fail→fail | 11,628 | 16,886 | +45% | 1 | 1 | 0% | 1,045 | 786 | -25% | 0 | 0 | — |
case-11 | fail→fail | 10,371 | 22,488 | +117% | 1 | 1 | 0% | 935 | 1,651 | +77% | 0 | 0 | — |
case-12 | fail→fail | 9,281 | 18,017 | +94% | 1 | 1 | 0% | 704 | 998 | +42% | 0 | 0 | — |
case-13 | fail→fail | 9,693 | 26,601 | +174% | 1 | 1 | 0% | 861 | 960 | +11% | 0 | 0 | — |
case-14 | fail→fail | 21,687 | 18,307 | -16% | 1 | 1 | 0% | 1,120 | 944 | -16% | 0 | 0 | — |
case-15 | fail→fail | 10,194 | 18,129 | +78% | 1 | 1 | 0% | 833 | 912 | +9% | 0 | 0 | — |
case-16 | fail→fail | 11,648 | 17,887 | +54% | 1 | 1 | 0% | 1,082 | 1,056 | -2% | 0 | 0 | — |
case-17 | fail→fail | 10,187 | 26,336 | +159% | 1 | 1 | 0% | 832 | 793 | -5% | 0 | 0 | — |
case-22 | pass→fail | 12,674 | 16,658 | +31% | 1 | 1 | 0% | 1,442 | 1,121 | -22% | 0 | 0 | — |
case-18 | fail→fail | 9,735 | 16,761 | +72% | 1 | 1 | 0% | 721 | 966 | +34% | 0 | 0 | — |
case-19 | fail→fail | 18,030 | 25,957 | +44% | 1 | 1 | 0% | 551 | 829 | +50% | 0 | 0 | — |
case-20 | fail→fail | 15,811 | 16,142 | +2% | 1 | 1 | 0% | 1,928 | 816 | -58% | 0 | 0 | — |
case-21 | fail→fail | 11,205 | 16,007 | +43% | 1 | 1 | 0% | 1,037 | 999 | -4% | 0 | 0 | — |
case-23 | pass→fail | 11,884 | 18,772 | +58% | 1 | 1 | 0% | 1,199 | 1,098 | -8% | 0 | 0 | — |
case-24 | fail→fail | 13,987 | 15,877 | +14% | 1 | 1 | 0% | 1,527 | 848 | -44% | 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. 24 cases were attempted, and 1 counted toward the lift figure. The other 23 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. A headline lift is not published for this run.
Other measured skills in the registry, with their headline benchmark lift.