copilot-data-readiness-protocol

---

name: copilot-data-readiness-protocol description: Use this skill before enabling Microsoft 365 Copilot for any user population. Runs an oversharing assessment, applies sensitivity labels and DLP controls, validates permissions baseline, and confirms a secure data foundation exists. Orchestrates m365-copilot-readiness-governance-agent as primary, m365-identity-zero-trust-agent for identity-layer controls, and copilot-governance-maestro-agent for cross-pillar sign-off. Hard refusal: Microsoft 365 Copilot must not be enabled for any user until the oversharing baseline is established and remediated. Never requests credentials, tenant IDs, or customer data. Production-impacting steps escalate to the relevant data owner or compliance team. allowed-tools: Read Grep Glob metadata: author: "github: Raishin" version: "0.1.0" updated: "2026-06-16" category: ai lifecycle: experimental

Copilot Data Readiness Protocol

Purpose

Microsoft 365 Copilot surfaces content that users already have permission to access. If that content is over-shared, unlabelled, or poorly governed, Copilot amplifies the exposure. This skill defines the mandatory readiness sequence that must complete before Copilot is enabled: oversharing baseline, sensitivity labelling, DLP guardrails, and a permissions baseline sign-off. It exists so organisations never enable Copilot into an ungoverned data estate. It does not configure production systems; it produces structured recommendations that data owners and security teams confirm.

When to use

• An organisation is planning to enable Microsoft 365 Copilot for the first time.
• Copilot is expanding to new user groups and the data estate for those users has not been assessed.
• A periodic data readiness review is due for an active Copilot deployment.
• An oversharing alert has been raised by Microsoft Purview DSPM and must be assessed before Copilot access continues.

When NOT to use

• Copilot is already enabled and this is a routine operational review with no new user groups — use the periodic governance review workflow instead.
• The matter is a data breach or security incident — route to the incident response protocol.
• The organisation has not yet deployed Microsoft Purview or SharePoint Advanced Management — prerequisites are not met; this protocol cannot proceed.
• You need live tenant configuration changes — escalate to the data owner; this protocol is recommendation-only.

Participating agents

• m365-copilot-readiness-governance-agent (primary — oversharing assessment, sensitivity labels, DLP, permissions baseline)
• m365-identity-zero-trust-agent (secondary — Conditional Access, identity-layer least privilege)
• copilot-governance-maestro-agent (cross-pillar sign-off and final readiness gate)

Inputs required

• Tenant-level SharePoint sharing settings and EEEU (Everyone except external users) status
• Microsoft Purview DSPM Data Risk Assessment output (or permission to run one)
• SharePoint Advanced Management (SAM) Content Management Assessment output
• List of user groups proposed for Copilot enablement
• Existing sensitivity label taxonomy and publishing scope
• DLP policy inventory for Microsoft 365 workloads

Evidence required

• Microsoft Purview is deployed and DSPM is active
• SharePoint Advanced Management is licensed and configured
• Sensitivity labels are published to users in scope
• Purview Audit is enabled for Copilot interaction activity
• A data owner is identified for each high-risk site

Workflow

1. Oversharing baseline — Run (or review) the Microsoft Purview DSPM Data Risk Assessment and SAM Content Management Assessment. Identify sites with EEEU access, broken permission inheritance, ownerless sites, inactive sites, and sensitive content exposed broadly.
2. Interim protections — Before remediating, apply SAM Restricted Content Discovery (RCD) to exclude high-risk sites from Copilot discovery. Apply Purview DLP policies scoped to the Copilot location to exclude sensitive content from grounding.
3. Sensitivity label review — Confirm that sensitivity labels are published to all users in scope. Identify content lacking labels on high-risk sites. Draft labelling recommendations for data owners.
4. Permissions remediation — For sites flagged as high-risk: remove excessive access and company-wide sharing links, correct broken inheritance, rescope sharing to approved users or groups, confirm site ownership.
5. Gate 1 — Oversharing remediation sign-off — Confirm that all critical-risk sites have been remediated or have accepted interim protections. Do not proceed to Copilot enablement without this sign-off from the data owner and copilot-governance-maestro-agent.
6. Identity layer check — Invoke m365-identity-zero-trust-agent to confirm that Conditional Access policies for Copilot users enforce MFA and device compliance. Flag any gaps.
7. DLP guardrails — Confirm DLP policies cover the Copilot location and are actively monitoring sensitive content interactions. Validate Insider Risk Management (IRM) adaptive protection is configured for high-risk users.
8. Gate 2 — Permissions baseline — Produce a permissions baseline document: confirmed label coverage rate, EEEU status, high-risk site count (open vs. remediated), DLP policy coverage. This is the minimum viable baseline before Copilot enablement.
9. Copilot readiness recommendation — copilot-governance-maestro-agent produces a go / conditional-go / no-go recommendation per user group. Conditional-go requires time-bound remediation commitments.
10. Human confirmation — Route the readiness recommendation to the data owner, security team, and Copilot programme owner for final sign-off. This protocol never enables Copilot autonomously.

Decision gates

Gate	Condition	Action
Oversharing baseline	Purview DSPM or SAM assessment not run	Block enablement; run assessment first
Critical-risk sites unmediated	High-risk sites without interim protection or remediation	Block enablement; apply RCD and DLP interim controls
EEEU active at tenant level	Everyone except external users enabled tenant-wide	Escalate to SharePoint admin; recommend disabling before enablement
Sensitivity label gap	>20% of content on high-risk sites lacks labels	Require labelling sprint before go-live
Identity layer gap	MFA or device compliance not enforced for Copilot users	Invoke m365-identity-zero-trust-agent; hold enablement

Refusal triggers

• Hard refusal: do not recommend Copilot enablement without an oversharing baseline. This is unconditional.
• Stop if the Purview DSPM assessment has not been run — do not substitute manual estimates for a real assessment.
• Stop if credentials, tenant IDs, or customer PII are requested to perform this assessment — refuse and escalate.
• Stop if the data owner cannot be identified for a critical-risk site — flag as a blocker; do not proceed past Gate 1.

Handoff rules

• All handoffs carry: tenant_scope, skill_id, skill_version, invoked_by, oversharing_risk_level, label_coverage_rate, gate_status, open_questions, do_not_do_list.
• Gate 1 sign-off must include the data owner name/role, the date, and a statement of residual risk accepted.
• copilot-governance-maestro-agent's readiness recommendation is the final cross-pillar artefact; it must not be bypassed.

KPIs

• Oversharing risk sites: critical / high / medium count (pre- vs. post-remediation)
• Sensitivity label coverage rate (% of files on high-risk sites labelled)
• EEEU status: tenant-level and site-level
• DLP policy coverage for Copilot location
• Time from assessment to enablement (days)

References

• https://learn.microsoft.com/microsoft-365/copilot/secure-govern-copilot-foundational-deployment-guidance
• https://learn.microsoft.com/microsoft-365/copilot/configure-secure-governed-data-foundation-microsoft-365-copilot
• https://learn.microsoft.com/sharepoint/get-ready-copilot-sharepoint-advanced-management
• https://learn.microsoft.com/purview/data-security-posture-management-learn-about
• https://learn.microsoft.com/security/zero-trust/copilots/zero-trust-microsoft-365-copilot