Viva Insights Sample Code

Workflow support

Workflow support

This table owns cross-workflow capability summaries. Workflow READMEs and references own exact input/configuration contracts and commands. Implementation support is not proof of a real-data trial or an executed agent evaluation.

Workflow	Supported path and runtime	Inputs and measures	Outputs and limitations	Verification entry point
Core analysis skill	Agent-guided R/Python using installed vivainsights APIs	Documented query exports with verified grain and metric semantics	Generated scripts, descriptive reports and readiness audits; not a universal export adapter or causal engine	Bundled examples/README.md and installed API inspection
Adoption prompts and starter	Copyable agent-guided R/Python workflow	Person Query, explicit eligibility for adoption denominators, verified activity and optional HR/assisted-hours metrics	Static HTML, memo, segmentation or editable deck; output code and numbers require verification; ROI is an assumption-based scenario	Prompt instructions and starter input/output specifications; validation metadata alone is not an executed test
Consumption dashboard runner	Maintained local R runner; synthetic demo plus scoped real-export build	Documented Consumption layouts, explicit column mappings, historical metadata and additive-credit decisions	Real builds are credits-only supported panels; no invented tokens, currency, Person Query adapter or ways-of-working panels	Runner README, tests/, and dashboard evals/README.md
GitHub developer dashboard	Synthetic R reproduction; real exports inspect/readiness only	Synthetic activity, Person Query and roster fixture; real headers inspected against the contract	No implemented real GitHub build adapter; no productivity or causal interpretation	Runner README and dashboard GitHub manifest
Person and Agent integration	Schema/join guidance, not a packaged dashboard adapter	Current selected-query join contract, historical identity and weekly Person spine	No implemented real Person/Agent dashboard adapters; documented joins do not extend runner support	Current join contract and its deterministic tests
Person Query causal skill	Agent-guided design and readiness; bounded synthetic Python/base-R runners	Verified exposure/outcome windows; fabricated balanced continuous-outcome panels in bundled runners	Synthetic readiness and limited linear comparisons, not general real-data PPML, DML, bootstrap or quasi-experimental execution	Causal examples/README.md and reference/evaluation.md; historical evidence is bounded there
Business-process collaboration skill	Agent-guided R/Python workflow; standard-library Python rehearsal and R/jsonlite classifier checks	Reviewed MQ scope/coverage, exact event decisions and optional purpose-specific PQ evidence	Descriptive contracts and generated analyses; bundled examples are fabricated. No general export adapter, process mining, inference engine or report renderer	Process skill examples/README.md, examples/test_examples.py and reference/evaluation.md
Copilot Causal Toolkit	Separate prescribed notebook/DML workflow	Selected toolkit notebook, its fixed treatment, dependencies and design-specific input contract	Do not replace with or claim equivalence to the synthetic causal runner; inspect identification and actual notebook preprocessing before causal interpretation	Toolkit prompt and selected notebook’s own diagnostics
External-collaboration integration	Standalone schemas/examples only where documented; no general cross-query adapter	Source-specific keys, grain and privacy must be established	No inferred identity bridge or assumed compatible join	Relevant source schema and example, not another query’s contract
Live product / MCP access	Unsupported by this repository	Local exported files only in maintained workflows	No MCP server, live-product connector or credentials workflow is supplied	Not applicable

Resolving references

For an installed core skill, examples/README.md is bundled locally. Optional workflow assets live in a retained sample-code checkout:

• Dashboard: frontier-analytics/starter-kits/copilot-query-dashboard/README.md, its tests/, and the dashboard skill’s report manifests and evals/README.md.
• Causal: the paired causal skill’s examples/README.md and reference/evaluation.md.
• Process: the paired process skill’s examples/README.md and reference/evaluation.md, installed with its core references.
• DML: frontier-analytics/prompts/copilot-adoption/copilot-causal-toolkit.md and the notebook it selects.
• Joins: current guide and downloadable contract.

Do not treat a missing optional checkout as missing core analysis functionality. Ask for the specific asset only when that workflow is requested. Keep the analysis, dashboard, process, causal and DML workflow boundaries explicit.

Catalogue summaries

These marked plain-text lists are authored here for generated skill cards.

• Agent-guided local R/Python analysis. Export layouts and installed APIs require verification for each workflow.
• Verify installed APIs and generated outputs. Built-in and synthetic examples are not real-export validation.
• Prefer enabled-user/enabled-days fields. Documented native licensed-only M365 Copilot nulls mean unlicensed. Failed joins remain unresolved. These licence semantics do not apply to unlicensed Copilot Chat or Consumption credits.
• Use max(organisation requirement, workflow floor): generic 5, joined-query/dashboard/causal 10 distinct people. Protect complementary cells and never publish individual rankings.

• Requires a retained sample-code checkout and R; reference demo rendering also requires Pandoc. Skill installation does not install dependencies.
• Consumption real builds support scoped credits-only panels. GitHub real data is inspect/readiness only; real Person/Agent dashboard adapters are not implemented.
• Reference reproductions are synthetic. Do not invent missing panels, tokens, currency, productivity, ROI or causal effects from activity.
• Use max(organisation requirement, 10 distinct people) across published and embedded output, protect complementary cells and never publish individual rankings.

• Guided design/readiness workflow, not a general real-data causal engine. Bundled R/Python runners use fabricated balanced panels and limited linear comparisons.
• Choose a supported identifying design and compatible diagnostics first; readiness-only and unavailable checks are valid. No general PPML, DML, bootstrap or quasi-experimental runner is implemented here.
• The prescribed Copilot Causal Toolkit remains a separate notebook/DML workflow; do not silently substitute treatments or methods.
• Install alongside the core analysis skill. Use max(organisation requirement, 10 distinct people), protect complementary cells and never publish individual rankings or person-specific effects.

• Agent-guided collaboration analysis around business processes using Meeting Query and optional purpose-specific Person Query exports.
• Requires approved title/date scope and verified coverage before reporting process totals. Purview audit logs and process mining are separate.
• Bundled Python rehearsal and R classifier checks use fabricated audit events; no general export adapter, inference engine or report renderer is implemented.
• Install alongside the core skill. Use max(organisation requirement, 10 distinct people), verify the applicable person base and keep event stability thresholds separate.

Back to Skills