kai-data-dashboard - Data To Dashboard Handoff
Kai root note: knowledge/, harness/, and scripts/ paths in this skill live in the Kai install, not the user's project. Resolve them against the first ancestor directory of this SKILL.md that contains a knowledge/ folder (the Kai plugin root, ~/.claude/kai, or the kai-cmo-harness repo). MARKETING.md, memory/, and any output files live in the current project. If a referenced scripts/ command is not available in this install, say so, skip it, and continue with the file-based guidance — never fabricate its output.
Turn existing Kai artifacts into a dashboard spec, data contract, or lightweight static dashboard. This skill is a companion surface. It should not replace analytics setup, audit analysis, or outbound strategy.
Use /kai-analytics when the user needs a tracking plan or attribution model. Use /kai-html-presentation when the user needs a client-ready audit deck. Use /kai-client-dashboard when the ask is a durable, white-labeled, client-facing product with its own onboarding wizard and page set — that skill calls this one for the data-contract layer underneath it. Use this skill when the data already exists and needs to become a dashboard-ready operator surface.
Phase 0: Identify The Source
Accept these inputs:
workspace/sdr-operator/<package-slug>/- Any folder with
kai-data.json, audit-data.json, _data-sources.md, or _data-gaps.md
- CSV exports from CRM, ESP, sequencer, ads, analytics, or sales tools
- Markdown reports with source-backed findings
- User-provided metrics and targets
If there is no source folder or file, ask for it. Do not fabricate sample data unless the output is explicitly labeled internal_demo.
Phase 1: Load Provenance
Before designing the dashboard:
- Read
_data-sources.md, _data-gaps.md, and available JSON/CSV files.
- Declare data mode:
sales_external, onboarding_connected, user_provided, or internal_demo.
- List unsupported fields as gaps.
- Do not add numbers that are not present in the source.
For audit folders, run:
python scripts/quality_gates/audit_provenance_lint.py <source-folder> --audit-dir
Phase 2: Choose Dashboard Type
Pick the dashboard type from the source and request:
| Type |
Best Fit |
Primary View |
sdr_operator_room |
SDR package, lead ledger, reply data |
Pipeline state, source quality, next actions |
marketing_ops |
Campaign, content, SEO, ad, lifecycle data |
Channel performance and bottlenecks |
executive_scorecard |
Monthly/weekly report |
KPIs, decisions, risks, next steps |
audit_delivery |
Audit folder |
Findings, scorecards, fixes, data gaps |
connector_health |
API sync or integration data |
Source freshness, failures, missing credentials |
Default to sdr_operator_room when the source is from /kai-sdr-operator.
Phase 3: Produce Dashboard Artifacts
Write output to:
<source-folder>/dashboard/
Required files:
dashboard-spec.md
metrics-dictionary.md
data-contract.json
source-map.md
data-gaps.md
Optional file when the user asks for a usable static artifact:
index.html
Do not build a full frontend app unless the user asks for implementation. For app builds, hand the spec to the relevant frontend skill or repository code.
Dashboard Spec Requirements
Each dashboard spec must include:
- Audience: executive, operator, SDR, marketer, client, founder, or analyst.
- Jobs to be done.
- Metric definitions with exact formulas.
- Data source per metric.
- Refresh cadence and freshness warning.
- Widgets, filters, drilldowns, empty states, and error states.
- Alert thresholds with source or hypothesis label.
- Permissions and sensitive-data handling.
- Handoff notes for frontend, BI, or static HTML build.
For SDR dashboards, include:
- Status counts by
sourced, enriched, approved_for_copy, queued, sent, replied, meeting_booked, disqualified, suppressed, and blocked.
- Source quality table.
- Fit score distribution.
- Next-action queue.
- Reply triage categories.
- Suppression, bounce, opt-out, and complaint warnings.
- Data gaps that block live outreach.
Static HTML Rules
If writing index.html:
- Keep it single-file unless the user requests app integration.
- Use tables for dense operator data.
- Use restrained styling, readable status colors, and responsive layouts.
- Keep critical numbers visible as text, not only canvas or images.
- Include a source footer or source drawer.
- Include empty states for missing metrics.
- Do not hide gaps. Show them as a first-class panel.
Quality Gates
Before handoff:
- Confirm every number has a source, retrieval date, or
internal_demo label.
- Confirm every metric has a formula or definition.
- Confirm
_data-gaps.md or data-gaps.md is represented.
- Confirm no placeholder text remains.
- Confirm sensitive fields are either excluded, masked, or explicitly approved.
- If HTML is produced, check desktop and mobile readability.
Output Summary
Final response should include:
- Dashboard folder path.
- Dashboard type.
- Files produced.
- Data gaps.
- Whether a static HTML dashboard was built or only specified.
1---2name: kai-data-dashboard3description: Convert Kai workflow data, CSV exports, audit folders, SDR package outputs, and marketing reports into dashboard-ready specs or lightweight static dashboards. Use when "data dashboard", "operator dashboard", "operator room", "HTML operators room", "sales dashboard", "SDR dashboard", "turn this data into a dashboard", "dashboard handoff", "visualize Kai data", or any request to package sourced marketing, sales, audit, or SDR data for a dashboard or presentation surface.4---56# kai-data-dashboard - Data To Dashboard Handoff78> **Kai root note:** `knowledge/`, `harness/`, and `scripts/` paths in this skill live in the Kai install, not the user's project. Resolve them against the first ancestor directory of this SKILL.md that contains a `knowledge/` folder (the Kai plugin root, `~/.claude/kai`, or the kai-cmo-harness repo). `MARKETING.md`, `memory/`, and any output files live in the current project. If a referenced `scripts/` command is not available in this install, say so, skip it, and continue with the file-based guidance — never fabricate its output.910Turn existing Kai artifacts into a dashboard spec, data contract, or lightweight static dashboard. This skill is a companion surface. It should not replace analytics setup, audit analysis, or outbound strategy.1112Use `/kai-analytics` when the user needs a tracking plan or attribution model. Use `/kai-html-presentation` when the user needs a client-ready audit deck. Use `/kai-client-dashboard` when the ask is a durable, white-labeled, client-facing product with its own onboarding wizard and page set — that skill calls this one for the data-contract layer underneath it. Use this skill when the data already exists and needs to become a dashboard-ready operator surface.1314## Phase 0: Identify The Source1516Accept these inputs:1718- `workspace/sdr-operator/<package-slug>/`19- Any folder with `kai-data.json`, `audit-data.json`, `_data-sources.md`, or `_data-gaps.md`20- CSV exports from CRM, ESP, sequencer, ads, analytics, or sales tools21- Markdown reports with source-backed findings22- User-provided metrics and targets2324If there is no source folder or file, ask for it. Do not fabricate sample data unless the output is explicitly labeled `internal_demo`.2526## Phase 1: Load Provenance2728Before designing the dashboard:29301. Read `_data-sources.md`, `_data-gaps.md`, and available JSON/CSV files.312. Declare data mode: `sales_external`, `onboarding_connected`, `user_provided`, or `internal_demo`.323. List unsupported fields as gaps.334. Do not add numbers that are not present in the source.3435For audit folders, run:3637```bash38python scripts/quality_gates/audit_provenance_lint.py <source-folder> --audit-dir39```4041## Phase 2: Choose Dashboard Type4243Pick the dashboard type from the source and request:4445| Type | Best Fit | Primary View |46|---|---|---|47| `sdr_operator_room` | SDR package, lead ledger, reply data | Pipeline state, source quality, next actions |48| `marketing_ops` | Campaign, content, SEO, ad, lifecycle data | Channel performance and bottlenecks |49| `executive_scorecard` | Monthly/weekly report | KPIs, decisions, risks, next steps |50| `audit_delivery` | Audit folder | Findings, scorecards, fixes, data gaps |51| `connector_health` | API sync or integration data | Source freshness, failures, missing credentials |5253Default to `sdr_operator_room` when the source is from `/kai-sdr-operator`.5455## Phase 3: Produce Dashboard Artifacts5657Write output to:5859```text60<source-folder>/dashboard/61```6263Required files:6465```text66dashboard-spec.md67metrics-dictionary.md68data-contract.json69source-map.md70data-gaps.md71```7273Optional file when the user asks for a usable static artifact:7475```text76index.html77```7879Do not build a full frontend app unless the user asks for implementation. For app builds, hand the spec to the relevant frontend skill or repository code.8081## Dashboard Spec Requirements8283Each dashboard spec must include:8485- Audience: executive, operator, SDR, marketer, client, founder, or analyst.86- Jobs to be done.87- Metric definitions with exact formulas.88- Data source per metric.89- Refresh cadence and freshness warning.90- Widgets, filters, drilldowns, empty states, and error states.91- Alert thresholds with source or hypothesis label.92- Permissions and sensitive-data handling.93- Handoff notes for frontend, BI, or static HTML build.9495For SDR dashboards, include:9697- Status counts by `sourced`, `enriched`, `approved_for_copy`, `queued`, `sent`, `replied`, `meeting_booked`, `disqualified`, `suppressed`, and `blocked`.98- Source quality table.99- Fit score distribution.100- Next-action queue.101- Reply triage categories.102- Suppression, bounce, opt-out, and complaint warnings.103- Data gaps that block live outreach.104105## Static HTML Rules106107If writing `index.html`:108109- Keep it single-file unless the user requests app integration.110- Use tables for dense operator data.111- Use restrained styling, readable status colors, and responsive layouts.112- Keep critical numbers visible as text, not only canvas or images.113- Include a source footer or source drawer.114- Include empty states for missing metrics.115- Do not hide gaps. Show them as a first-class panel.116117## Quality Gates118119Before handoff:1201211. Confirm every number has a source, retrieval date, or `internal_demo` label.1222. Confirm every metric has a formula or definition.1233. Confirm `_data-gaps.md` or `data-gaps.md` is represented.1244. Confirm no placeholder text remains.1255. Confirm sensitive fields are either excluded, masked, or explicitly approved.1266. If HTML is produced, check desktop and mobile readability.127128## Output Summary129130Final response should include:131132- Dashboard folder path.133- Dashboard type.134- Files produced.135- Data gaps.136- Whether a static HTML dashboard was built or only specified.
cgallic