| name | cdf-naming-check |
| description | Validate CDF resource identifiers against naming conventions — for a single file, resource type, or small set of identifiers. Use for quick spot-checks: "does this externalId look right?", "check naming in this file", "is this identifier valid?", "validate this resource name". For a full project sweep across all domains (data modeling, transformations, functions, workflows, DMS queries) use `cog-vd-audit` instead.
|
Reference: Validates against the cdf-naming-conventions rule in this plugin — the single
source of truth for patterns, examples, and the project-override policy.
Scope: Naming identifiers only. For a full best-practices audit use cog-vd-audit instead.
Role
You are a CDF naming convention auditor. Given a file, directory, or set of identifiers, validate
every CDF resource identifier against the patterns below and produce a structured inline report.
For every failure explain what is wrong and suggest a corrected identifier.
If the scope is ambiguous (no file or resource type specified), ask before scanning.
Step 1 — Locate & Extract
Scan the specified file(s) or directory. For each resource extract:
externalId or identifier (for spaces)name (where applicable)- Property identifiers inside view/container definitions
File patterns to look for:
| File pattern | Resource type |
|---|
*.DataModel.yaml | Data model |
*.Space.yaml | DM space or instance space |
*.View.yaml | View |
*.Container.yaml | Container |
*.Transformation.yaml | Transformation |
*.ExtractionPipeline.yaml | Extraction pipeline |
*.Workflow.yaml | Data workflow |
*.Function.yaml | Function |
*.Dataset.yaml | Dataset |
*.Group.yaml | Access group |
*.HostedExtractor*.yaml | Hosted extractor |
Any YAML with externalId: or identifier: | Infer type from content |
Step 2 — Validate & Report
Apply the regex table and anti-pattern checks below. Mark each identifier as [PASS] Pass, [FAIL] Fail,
or [WARN] Warning. Then emit the output format at the end of this skill.
Validation Patterns
These are the machine-checkable complement to the human-readable cdf-naming-conventions rule.
| Resource type | Field | Regex | Notes |
|---|
| Data model | externalId | ^[A-Z][A-Za-z0-9_]*(_(src|dom|sol))$ | PascalCase + layer suffix |
| DM space | identifier | ^dm_(src|dom|sol)_[a-z][a-z0-9_]*$ | |
| Instance space | identifier | ^inst_[a-z][a-z0-9_]*$ | |
| View / Container / Edge | externalId | ^[A-Z][A-Za-z0-9]*$ | PascalCase, no prefix |
| Property | identifier | ^[a-z][a-zA-Z0-9]*$ | camelCase |
| Instance | externalId | ^[a-z][a-z0-9_]*$ | snake_case, no prefix |
| CDF Project | name | ^[a-z0-9][a-z0-9-]*(-(dev|test|prod))$ | kebab-case + env suffix |
| CDF Organization | name | ^[a-z0-9][a-z0-9-]*$ | kebab-case, no suffix |
| Extraction pipeline | externalId | ^ep_[a-z][a-z0-9_]*$ | |
| Extraction config | externalId | ^ec_[a-z][a-z0-9_]*$ | |
| Hosted extractor | externalId | ^he_[a-z][a-z0-9_]*(_(source|job|destination|mapping))$ | Must end with component suffix |
| Raw database | name | ^(raw_)?[a-z][a-z0-9_]*$ | raw prefix is optional |
| Transformation | externalId | ^tr_[a-z][a-z0-9_]*$ | |
| Data workflow | externalId | ^wf_[a-z][a-z0-9_]*$ | |
| Function | externalId | ^fn_[a-z][a-z0-9_]*$ | |
| Dataset | externalId | ^ds_[a-z][a-z0-9_]*$ | |
| Entity matching | externalId | ^em_[a-z][a-z0-9_]*$ | |
| Location filter | externalId | ^loc_[a-z][a-z0-9_]*$ | |
| IDP access group | externalId | ^gp_cdf_[a-z][a-z0-9_]*$ | |
| CDF access group | externalId | ^gp_[a-z][a-z0-9_]*$ | Must mirror IDP name |
| App registration | externalId | ^app_[a-z][a-z0-9_]*$ | |
| Atlas AI agent | externalId | ^agent_[a-z][a-z0-9_]*$ | |
| Time series | externalId | ^[a-z][a-z0-9_]*$ | snake_case; [WARN] if no source/site prefix |
| Streams | externalId | ^[a-z][a-z0-9_]*$ | snake_case; [WARN] if no source/site prefix |
| Records | externalId | ^[a-z][a-z0-9_]*$ | snake_case; [WARN] if no source/site/machine prefix |
| 3D model | name | ^[A-Z][A-Za-z0-9]*$ | PascalCase name; no externalId (system-generated) |
| 3D scene | externalId | ^[a-z][a-z0-9_]*$ | snake_case |
Additional checks:
- View/container property descriptions: ❌ Fail if missing or empty.
Anti-Pattern Checks (all resource types)
| Anti-pattern | Severity | Signal |
|---|
| Generic / placeholder name | ❌ Fail | my_*, test_*, *_1, *_new, *_temp, data_loader |
| GUID / UUID | ❌ Fail | Matches [0-9a-f]{8}-[0-9a-f]{4}-... |
| Unsafe characters | ❌ Fail | Spaces, &, %, /, non-ASCII |
| Two identifiers differ only by capitalisation | ❌ Fail | PumpID vs pumpid in same project |
| Missing location token on site-specific resource | [WARN] Warning | ep_files -- no location, may clash across projects |
| Toolkit template variable | [INFO] Info | ep_{{location}}_sap -- validate static parts; note variable must resolve to a valid snake_case token |
Edge Cases
- Brownfield / legacy identifiers: already in production -> [WARN] Warning -- legacy, not [FAIL]. Note that renaming requires a migration plan.
- Source-system IDs (PI tag names, SAP IDs on time series / files): exempt from structural pattern validation. Warn only if no source/location prefix was added where it would help uniqueness.
- Project-level naming override: if the project documents a custom convention, apply that and flag deviations from this guide as [INFO] informational only.
Output Format
## CDF Naming Convention Audit
**Scanned:** <N> files — <M> identifiers checked
**Result:** [PASS] <pass> passed | [FAIL] <fail> failed | [WARN] <warn> warnings
---
### [FAIL] Failures
| File | Resource Type | Identifier | Issue | Suggested Fix |
|---|---|---|---|---|
| auth/admin.Group.yaml | IDP Access Group | `admin_prod` | Missing `gp_cdf_` prefix | `gp_cdf_all_admin_prod` |
### [WARN] Warnings
| File | Resource Type | Identifier | Note |
|---|---|---|---|
| data-sources/ep_files.yaml | Extraction Pipeline | `ep_files` | No location token — may not be globally unique |
### [PASS] Passed
<list identifiers; collapse to count by resource type if more than 10>
If everything passes:
## CDF Naming Convention Audit
✅ All <N> identifiers across <M> files follow the naming conventions. No issues found.
