mcpField
E
experimentation-platform-governance-review

experimentation-platform-governance-review

github.com/sylphxaiData & databases140 installsUpdated today
Repository

About experimentation-platform-governance-review

This skill turns experimentation governance questions into review artifacts with owners, gates, metrics, and recovery paths. It is for teams planning or assessing A/B tests and other experiments, with attention to risk, conflicts, ethics, and decision records.

Key features

  1. 01Identify experiment type, hypothesis, and risk
  2. 02Define eligibility, randomization, and metrics
  3. 03Set conflict policies and ethics review
  4. 04Generate governance plans and decision tables
  5. 05Create checklists and review cadence

Use cases

  1. 01Review risks in pricing or ranking experiments
  2. 02Plan conflict handling for overlapping tests
  3. 03Set guardrail metrics and decision thresholds

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/sylphxai/skills --skill experimentation-platform-governance-review

After installing, type/experimentation-platform-governance-reviewin Claude Code to confirm it loaded.

Bundled filesagents/openai.yamlreferences/experimentation-platform-governance-patterns.md

Details

From the project README

Experimentation Platform Governance Review

Use this skill to convert a experimentation governance, A/B tests, holdouts, metrics, guardrails, exposure conflicts, ethics, and decision review question into a concrete artifact with owners, gates, metrics, and recovery paths.

Workflow

  1. Identify experiment type, hypothesis, exposure unit, target population, metrics, risk class, overlapping tests, owner, and decision deadline.
  2. Read references/experimentation-platform-governance-patterns.md.
  3. Classify experiment as UX, pricing, ranking, lifecycle, onboarding, game economy, AI, permission, or infrastructure/product-risk test.
  4. Define eligibility, randomization, holdouts, metrics, guardrails, sample plan, conflict policy, ethics review, launch gate, and decision record.
  5. Produce governance plan, state machine, decision table, event schema, checklist, and experiment review cadence.

When not to use

  • Do not use for generic advice the base model already handles without this skill's specific artifact contract.

Guardrails

  • Do not test pricing, ranking, safety, privacy, or accessibility changes without risk review and support/comms plan.
  • Do not run overlapping experiments that contaminate metrics without conflict policy.
  • Do not declare winners without guardrails, segment review, and decision threshold.
  • Do not keep long-lived holdouts or experiments without user-impact and cleanup ownership.

Output format

Experiment context:
Hypothesis / exposure / risk class / conflicts:

Experiment governance plan:
| Area | Decision | Evidence | Risk | Owner |
| --- | --- | --- | --- | --- |

Metrics, guardrails, and conflicts:
- <item> -> <policy, metric, edge case, support note>

Decision and cleanup policy:
- <trigger> -> <action, communication, owner>

Comments

Markdown supported