OpenParable
Skip to content
← Back to skills

Asim Parser Filter Validator

Testing & Quality
asim-parser-filter-validator

Asim Parser Filter Validator is an Agent Skill for testing & quality published in the Azure/Azure-Sentinel repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Validates ASIM parser filtering parameters by running filter tests against a Log Analytics workspace. Pure PowerShell using az CLI for authentication — no Python or Azure SDK packages required. Use this skill after creating or modifying an ASIM parser to verify that its filtering parameters work correctly.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/Azure/Azure-Sentinel/blob/HEAD/.github/skills/asim-parser-filter-validator/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/asim-parser-filter-validator, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

ASIM Parser Filter Validator

Validates that an ASIM parser's filtering parameters (e.g. disabled, starttime, endtime, srcipaddr_has_any_prefix, etc.) behave correctly by running queries against a Log Analytics workspace.

This skill is a pure PowerShell implementation that uses az CLI for authentication. No Python, Azure SDK packages, or YAML modules are required.

  • PowerShell 7+
  • Azure CLI (logged in via az login)

What it tests

For every filtering parameter declared in the parser's KQL function signature:

Parameter typeTests performed
disableddisabled=true returns 0 rows; disabled=false returns rows
datetime (starttime / endtime)Filtering by the midpoint timestamp returns fewer rows than unfiltered
Scalar (string, int)Filtering by a real value returns exactly that value; filtering by a fictitious value returns 0 rows
dynamic (*_has_any, *_has_all, *_has_any_prefix)Filtering by one/two real values returns fewer rows; fictitious value returns 0 rows; substring / prefix variants tested where applicable

Supported schemas

AgentEvent, AlertEvent, AssetEntity, AuditEvent, Authentication, DhcpEvent, Dns, FileEvent, NetworkSession, ProcessEvent, RegistryEvent, UserManagement, WebSession.

Inputs

InputRequiredDescription
Parser KQL pathYesPath to the ASIM parser .kql file to test
Schema nameYesASIM schema name (e.g. Dns, Authentication, NetworkSession)
Workspace IDYesLog Analytics workspace GUID to run queries against

How to run

.\.github\skills\asim-parser-filter-validator\scripts\asimFilterTest.ps1 -ParserFile "{PathToFilterParserKQL}" -SchemaName "{SchemaName}" -WorkspaceId "{your-workspace-guid}"

Prerequisites

  1. Azure CLI authenticated — run az login if not already logged in. If you get an authentication error, use the az-cli-command-runner skill to verify login status.
  2. The workspace must contain data for the tables referenced by the parser.

Interpreting results

  • Green — test passed.
  • FAIL — the filtering parameter did not behave as expected. The failure message indicates which parameter failed and what was expected.
  • Known partial validations — AuditEvent (EventResult), Authentication (EventType), and Dns (EventType) have known single-failure scenarios that are automatically ignored.

Known acceptable failure reasons (can be ignored)

When analyzing filter-validation failures, a failure can be ignored if it matches one of these inherent-limitation categories:

  1. Documented schema known exception — The schema defines a known exception for a parameter (e.g. a multi-value test that cannot succeed because the value combinations do not co-exist in that schema's data). Check whether the ASIM schema documentation lists the parameter as a known exception.
  2. Constant or single-value field by design — The parser maps the filtered field to a single constant value, so multi-value or reduction-based filtering cannot reduce rows further. This is expected when the source provides only one possible value for that field.
  3. Insufficient distinct values in the test window — The 2-day query window does not contain enough distinct values for the field to exercise multi-value filtering. This is a data-availability limitation, not a parser bug.

When reviewing filter-validation output, analyze each failure message against these categories. If all other parameters pass and every failure maps to one of the reasons above, the parser's filtering implementation is considered correct.

Troubleshooting

SymptomFix
Failed to acquire access token via az CLIRun az login
No data in the provided workspaceEnsure the workspace has ingested data for the relevant tables within the last 2 days
Schema: X - Not supportedThe schema name is not in the supported list — update the $AllSchemasParameters hashtable in asimFilterTest.ps1