Install any skill in seconds. Free to start, no credit card required.
Get Started Free →OverKill Hill P³ SharePoint List build screenshot auditor. Use when reviewing screenshots of SharePoint Online list settings, column settings, indexed columns, content types, views, or Advanced Settings against a repo schema, build card, runbook, ADR, or CSV header. Also activate when the user says they captured SharePoint build screenshots, asks whether a list or column was configured correctly, points to arrows/markups in images, or needs a defect list before continuing manual SharePoint build
.claude/skills/okhp3-sp-build-auditor/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✗→✓ | ▲ Improved | 51% | 0% |
| case-20 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 41% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 9% | 0% |
OverKill Hill P³ · overkillhill.com · github.com/OKHP3
Use this skill to audit SharePoint UI screenshots captured during a manual list build. It turns screenshots into actionable build findings by comparing the visible UI settings against the repo's build cards, schema, and ADR decisions.
| In scope | Out of scope | |----------|-------------| | Screenshot review for list settings, column settings, indexes, views, and content types | Making live SharePoint changes | | Matching UI evidence to schema/v2-build-cards.md | Replacing human site-owner approval | | Finding sticky-name, type, required, hidden, index, and delete-behavior issues | Optical perfection or image cleanup | | Producing build corrections and v2-delta.md notes | Full data migration validation |
schema/v2-build-cards.md.schema/schema-fields.md if the card is ambiguous.matches build card, must fix before data load, should fix before broad use, acceptable deviation if logged, or needs another screenshot.v2-delta.md note. If it only corrects execution, recommend a SharePoint UI fix.Choice columns (single-select or multi-select) cannot be indexed in SharePoint Online. The Edit Index UI does not offer Choice columns in the Primary Column dropdown. Any build card or plan that lists a Choice column as a planned index is incorrect and must be corrected before build.
Indexable types: Single line of text, Number, Date and Time, Yes/No, Lookup, Person/Group, Managed Metadata. Non-indexable types: Choice, Multiple lines of text, Hyperlink, Calculated, Currency.
Impact on auditing: When reviewing an Indexed Columns screenshot, do not flag the absence of a Choice column as a missing index. Flag it as a build card correction required. If the builder created an index on an alternative non-Choice column instead, evaluate whether that substitution is documented and intentional.
For list-level screenshots:
For column screenshots:
| Severity | Use when | |---|---| | P0 stop build | Sticky URL/internal name is wrong, wrong list created, or wrong column exists that would require rebuild. | | P1 fix before data load | Column type, multi-choice vs single-choice, required setting, lookup behavior, index, or attachment setting conflicts with schema. | | P2 fix before broad use | Governance setting weakens reliability but does not corrupt data immediately, such as Quick Edit or offline sync. | | P3 document | Acceptable deviation, empty descriptions, or useful extra index that should be logged. |
From the analysis/artifacts/build-screenshots/01_ReferenceValues/ set reviewed in session sp-build-findings.md:
| Screenshot | Finding | Severity | Action | |---|---|---|---| | ReferenceGroup.png | Display choices using = Checkboxes (multi-select). Expected: Drop-Down Menu (single-select). | P1 fix before data load | Edit Column → Display choices using: Drop-Down Menu; Fill-in: No | | SourceSystem.png | Display choices using = Checkboxes (multi-select). Expected: Drop-Down Menu. | P1 fix before data load | Edit Column → Display choices using: Drop-Down Menu; Fill-in: No | | SortOrder.png | Number of decimal places = Automatic. Expected: 0. | P1 fix before data load | Edit Column → Decimal places: 0 | | ValueLabel.png | Description field empty. Max chars = 255 (correct for Single line of text). | P3 document | Add description text to column | | Definition.png | Text type = Plain text (correct). Description empty. | P3 document | Add description text | | MapsFromValues.png | Text type = Plain text (correct). Description empty. | P3 document | Add description text | | adv-01.png | Attachments = Enabled. Expected: Disabled. | P1 fix before data load | List Settings → Advanced Settings → Attachments: Disabled | | adv-02.png | Automatic Index Management = Yes. Expected: No for lists with manual index plan. | P1 fix before data load | List Settings → Advanced Settings → AIM: No | | indexes.png | Index being created on Reference Code (Single line of text). Build card listed ReferenceGroup as an index — but ReferenceGroup is a Choice column and cannot be indexed. The builder discovered this at build time. ReferenceCode is a valid substitute. IsActive should also be indexed. | Build card correction required | Remove ReferenceGroup from Card 01 index list; add ReferenceCode as deliberate index substitute; confirm IsActive is indexed | | content-type.png | Content type management appears enabled. Single content type "Item" present. IsActive shows Optional. | P3 document | IsActive is required per build card — verify required setting is confirmed in Edit Column, not only in content type view |
Read references/shot-review.md when reviewing a new screenshot folder or when the user asks for image-by-image findings.
Built by Jamie Hill · OverKill Hill P³ Published at github.com/OKHP3 Part of the OKHP3/skillz Agent Skill library. MIT License -- free to use, fork, and adapt. A nod to the source is appreciated.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | fail→pass | 10,205 | 6,222 | -39% | 1 | 1 | 0% | 1,745 | 2,635 | +51% | 0 | 0 | — |
case-20 | fail→pass | 12,362 | 11,314 | -8% | 1 | 1 | 0% | 2,292 | 2,999 | +31% | 0 | 0 | — |
case-01 | fail→pass | 16,408 | 15,709 | -4% | 1 | 1 | 0% | 3,004 | 4,771 | +59% | 0 | 0 | — |
case-02 | fail→pass | 20,589 | 18,527 | -10% | 1 | 1 | 0% | 3,861 | 5,438 | +41% | 0 | 0 | — |
case-03 | fail→pass | 25,167 | 23,704 | -6% | 1 | 1 | 0% | 4,915 | 5,370 | +9% | 0 | 0 | — |
case-04 | fail→pass | 9,641 | 4,146 | -57% | 1 | 1 | 0% | 1,490 | 2,345 | +57% | 0 | 0 | — |
case-05 | fail→pass | 9,335 | 3,464 | -63% | 1 | 1 | 0% | 1,499 | 2,239 | +49% | 0 | 0 | — |
case-06 | fail→pass | 9,878 | 3,691 | -63% | 1 | 1 | 0% | 1,502 | 2,285 | +52% | 0 | 0 | — |
case-07 | fail→pass | 4,189 | 3,030 | -28% | 1 | 1 | 0% | 646 | 2,081 | +222% | 0 | 0 | — |
case-08 | fail→pass | 7,390 | 3,798 | -49% | 1 | 1 | 0% | 1,111 | 2,353 | +112% | 0 | 0 | — |
case-09 | pass→pass | 4,525 | 2,749 | -39% | 1 | 1 | 0% | 767 | 2,120 | +176% | 0 | 0 | — |
case-10 | pass→pass | 7,605 | 3,581 | -53% | 1 | 1 | 0% | 1,367 | 2,253 | +65% | 0 | 0 | — |
case-11 | fail→pass | 3,467 | 3,545 | +2% | 1 | 1 | 0% | 433 | 2,199 | +408% | 0 | 0 | — |
case-12 | pass→pass | 7,355 | 1,828 | -75% | 1 | 1 | 0% | 1,070 | 2,019 | +89% | 0 | 0 | — |
case-13 | fail→pass | 5,813 | 3,094 | -47% | 1 | 1 | 0% | 895 | 2,068 | +131% | 0 | 0 | — |
case-15 | fail→pass | 8,915 | 5,862 | -34% | 1 | 1 | 0% | 1,368 | 2,213 | +62% | 0 | 0 | — |
case-16 | fail→pass | 11,133 | 10,488 | -6% | 1 | 1 | 0% | 1,548 | 3,376 | +118% | 0 | 0 | — |
case-17 | pass→pass | 5,795 | 4,926 | -15% | 1 | 1 | 0% | 927 | 2,462 | +166% | 0 | 0 | — |
case-18 | fail→pass | 8,967 | 1,695 | -81% | 1 | 1 | 0% | 1,339 | 1,862 | +39% | 0 | 0 | — |
case-19 | fail→pass | 14,346 | 8,400 | -41% | 1 | 1 | 0% | 2,192 | 2,909 | +33% | 0 | 0 | — |
case-21 | fail→pass | 7,642 | 3,027 | -60% | 1 | 1 | 0% | 1,317 | 2,112 | +60% | 0 | 0 | — |
case-22 | fail→pass | 14,606 | 5,596 | -62% | 1 | 1 | 0% | 2,852 | 2,599 | -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. 22 cases were attempted. The headline lift of +82 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.