agentleFS
PatrickGallucciPatrickGallucci/fabric-skills/skills/fabric-lakehouse-access-control/SKILL.md

fabric-lakehouse-access-control

Skill17 starsChanged 9 months agoMITView on GitHub

Troubleshoot Microsoft Fabric Lakehouse access control issues including OneLake security roles, SQL analytics endpoint permissions, workspace roles, data access roles, row-level security (RLS), column-level security (CLS), object-level security (OLS), dynamic data masking, shortcut permissions, Direct Lake security integration, DefaultReader role, ReadAll permission, and OneLake data access role conflicts. Use when users report permission denied, unauthorized access, missing data, empty query results, or cannot see tables in Fabric Lakehouse.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Follow this sequence to isolate the root cause.

    line 54
  2. Distinctive:

    Run the SQL permissions audit to enumerate explicit grants on a SQL analytics endpoint.

    line 86
  3. Distinctive:

    Run the access control checklist to validate a user's effective permissions across all security layers.

    line 88
  4. Distinctive:

    Use the RLS validation template as a starting point for testing row-level security predicates.

    line 90

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Microsoft Fabric Lakehouse Access Control remediate

Diagnose and resolve access control issues across all security layers of Microsoft Fabric Lakehouse, including workspace roles, item permissions, OneLake security data access roles, SQL analytics endpoint granular permissions, and Direct Lake semantic model security integration.

When to Use This Skill

  • User cannot access lakehouse data or sees "permission denied" errors
  • Queries return empty results or missing rows/columns unexpectedly
  • OneLake security roles not applying or taking too long to propagate
  • SQL analytics endpoint permissions not restricting access as expected
  • Direct Lake semantic model returns errors after enabling OneLake security
  • Shortcut data is inaccessible or returns 404 errors
  • DefaultReader role conflicts with custom data access roles
  • Deployment pipeline or git integration issues with data access roles
  • Users see data they should not have access to (over-provisioned)
  • Row-level or column-level security not filtering correctly

Prerequisites

  • Microsoft Fabric workspace with capacity or trial license
  • Admin, Member, or Contributor workspace role for managing security
  • Familiarity with T-SQL for SQL analytics endpoint security
  • PowerShell 7+ for running diagnostic scripts

Fabric Lakehouse Security Layers

Fabric Lakehouse enforces security through multiple overlapping layers. Understanding the evaluation order is critical for remediate.

Layer 1 - Workspace Roles control coarse access to all items in a workspace.

RoleWorkspace AccessOneLake Read/WriteManage Security
AdminFullYesYes
MemberFullYesYes
ContributorFullYesNo
ViewerSee items onlyNo (unless granted)No

Layer 2 - Item Permissions grant access to a single lakehouse without a workspace role. Key permissions: Read (see item), ReadAll (read all OneLake data), Reshare, Write.

Layer 3 - OneLake Security (Preview) provides granular role-based access at the table, folder, row, and column level. Enforced across all compute engines. Disabled by default; once enabled, cannot be turned off.

Layer 4 - SQL Analytics Endpoint Security uses T-SQL GRANT/DENY/REVOKE for object-level, row-level, and column-level security. Only applies to queries via the SQL analytics endpoint.

Layer 5 - Semantic Model Security applies RLS/OLS within Direct Lake or DirectQuery models only.

Quick Diagnostic Workflow

Follow this sequence to isolate the root cause. See workflow-diagnostics.md for full details.

TODO

  • Step 1: Identify the access path — see workflow-diagnostics.md
  • Step 2: Verify workspace role and item permissions — see workflow-diagnostics.md
  • Step 3: Check OneLake security role membership — see workflow-diagnostics.md
  • Step 4: Audit SQL analytics endpoint permissions — see workflow-diagnostics.md
  • Step 5: Validate Direct Lake security integration — see workflow-diagnostics.md
  • Step 6: Check shortcut and cross-region access — see workflow-diagnostics.md

Common Issues and Resolutions

SymptomLikely CauseResolution
User sees all data despite OneLake security rolesDefaultReader role still activeDelete or edit DefaultReader role to remove ReadAll grant
Empty results from queriesRLS filtering all rows for userCheck RLS predicates; verify user is member of correct role
"Can't find table" or "Column can't be found" in Direct LakeOLS/CLS hiding objects after OneLake security appliedAdd user to role with table/column access
Shortcut returns 404Cross-region shortcut with OneLake security enabledOneLake security does not support cross-region shortcuts
Permission changes not taking effectPropagation latencyRole definition changes: ~5 min. Group membership changes: ~1 hour. Some engines add another hour of caching
SQL endpoint shows data OneLake security should restrictSQL security and OneLake security are separate layersApply security at both layers or restrict access to only the secured endpoint
B2B guest user cannot access data via OneLake security roleEntra External ID settingsSet Guest user access to "same access as members" in Entra External Collaboration settings
Distribution list members can't access SQL endpointSQL endpoint can't resolve DL membershipUse security groups instead of distribution lists for OneLake security roles
Deployment pipeline fails for data access rolesTarget workspace DAR tracking not enabledManually enable DAR tracking on target workspace before deploying

OneLake Security Limitations

See onelake-security-reference.md for complete details.

Key limits to remember: maximum 250 roles per lakehouse, 500 members per role, 500 permissions per role. Mixed-mode queries accessing both OneLake-security-enabled and non-enabled data will fail. OneLake security does not work with private link protection or Azure/Purview Data Share.

Available Scripts

Run the SQL permissions audit to enumerate explicit grants on a SQL analytics endpoint.

Run the access control checklist to validate a user's effective permissions across all security layers.

Use the RLS validation template as a starting point for testing row-level security predicates.

References

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.