Alibabacloud Pai Feature Store Featuredb Usage Query
Query FeatureDB read/write usage data from PAI-FeatureStore. Use for analyzing consumption, usage trends, and cost estimation by official pricing. Supports date range queries with breakdowns.
Triggers: "FeatureDB read/write volume", "PAI-FeatureStore usage", "FeatureDB usage", "FeatureDB statistics", "Feature View usage", "PAI-FeatureStore billing"
Run aliyun version to verify >= 3.3.3. If not installed or version too low,
run curl -fsSL https://aliyuncli.alicdn.com/setup.sh | bash to update,
or see references/cli-installation-guide.md for installation instructions.
Then [MUST] run aliyun configure set --auto-plugin-install true to enable automatic plugin installation.
[MUST] run aliyun plugin update to ensure that any existing plugins on your local machine are always up-to-date.
[MUST] CLI User-Agent — Every aliyun CLI command invocation must include:
--user-agent AlibabaCloud-Agent-Skills/alibabacloud-pai-feature-store-featuredb-usage-query
The PAI-FeatureStore CLI plugin will be automatically installed when first used if auto-plugin-install is enabled.
Timeout: Default 10s is sufficient for all API calls.
Authentication
Pre-check: Alibaba Cloud Credentials Required
Security Rules:
NEVER read, echo, or print AK/SK values (e.g., echo $ALIBABA_CLOUD_ACCESS_KEY_ID is FORBIDDEN)
NEVER ask the user to input AK/SK directly in the conversation or command line
NEVER use aliyun configure set with literal credential values
NEVER output any string resembling a key (e.g., LTAI5t***, AccessKeyId: xxx, ak-***)
ONLY use aliyun configure list to check credential status
bash
aliyun configure list
Check the output for a valid profile (AK, STS, or OAuth identity).
If no valid profile exists, STOP here.
Required Parameters
IMPORTANT: Parameter Confirmation — Before executing any command or API call, ALL user-customizable parameters (e.g., RegionId, instance names, CIDR blocks, passwords, domain names, resource specifications, etc.) MUST be confirmed with the user. Do NOT assume or use default values without explicit user approval.
Parameter Name
Required/Optional
Description
Default Value
RegionId
Required
Alibaba Cloud region ID. Must be one of the supported regions listed above.
None
WorkspaceId
Optional*
PAI workspace ID (numeric, e.g., 12345). Required if DatasourceId is not provided.
None
DatasourceId
Optional*
FeatureDB datasource ID (numeric, e.g., 67890). Required if WorkspaceId is not provided.
None
StartDate
Optional
Query start date in yyyy-MM-dd format.
30 days ago (for total usage) or 7 days ago (for project/feature view)
EndDate
Optional
Query end date in yyyy-MM-dd format.
Yesterday
ProjectName
Optional
Project name (required for project-level and feature view queries).
None
FeatureViewName
Optional
Feature view name (required for feature view trend queries).
None
SortBy
Optional
Sort field: ReadCount or WriteCount
ReadCount
Order
Optional
Sort order: ASC or DESC
DESC
Note: Either WorkspaceId or DatasourceId must be provided. If neither is provided, the skill will list all available FeatureDB datasources for user selection.
RAM Permissions
This skill requires the following RAM permissions. See references/ram-policies.md for detailed policy configuration.
Required Actions:
paifeaturestore:ListInstances
paifeaturestore:GetDatasource
paifeaturestore:ListDatasources
paifeaturestore:ListDatasourceFeatureViews
Core Workflow
Functional Scope (MUST CHECK FIRST)
This skill ONLY supports querying FeatureDB read/write usage statistics. Any requests outside of "querying historical read/write counts or costs" are out of scope and must be politely declined with a list of supported features.
[MUST] If request is out of scope, reject politely and list supported features:
"Sorry, this Skill only supports querying FeatureDB read/write usage statistics, not [user request]. Supported features: 1) Daily read/write totals 2) Feature view details 3) Usage trends 4) Cost calculation"
Feature Selection Logic
Which features to execute depends on user's request:
Scenario
Action
Request out of scope or unsupported
Reject and list supported features (see above)
User specifies supported feature(s)
Execute the specified feature(s)
User doesn't specify any feature
Execute default behavior (see below)
Supported Functions
Query daily FeatureDB read/write totals from StartDate to EndDate with analysis (max 30-day span)
Query per-feature-view read/write counts for a specific date, with sorting and analysis
Query daily FeatureDB read/write totals for a specific project from StartDate to EndDate with analysis (max 30-day span)
Query daily FeatureDB read/write trends for a specific feature view from StartDate to EndDate with analysis (max 30-day span)
Calculate costs based on official pricing and query results
User-Friendly Prompts
Use descriptive names instead of function numbers when prompting, because "Function 3" is meaningless to users who haven't read this document.
[MUST] When user provides DatasourceId, you MUST call GetDatasource or ListDatasources to verify the datasource type is FeatureDB. Do NOT skip this validation step.
Validation: Check the Type field in the response. If Type is NOT FeatureDB, stop and inform the user that this datasource is not a FeatureDB datasource.
Case C: Neither WorkspaceId nor DatasourceId provided
Must select ONE datasource. These queries are datasource-specific and cannot span multiple datasources.
Error Handling:
If WorkspaceId or DatasourceId is invalid, fall back to Case C and list all available datasources.
Step 3: Query Usage Statistics
Date Validation (before API call):
Format: Must be yyyy-MM-dd (e.g., 2024-03-15). If user provides other formats (e.g., 3/15, 2024.03.15), convert to correct format.
Range: EndDate - StartDate must be ≤ 30 days. If exceeded, ask user to narrow the range or split into multiple queries.
Boundary: EndDate ≤ today; StartDate ≤ EndDate.
Relative Date Ranges ("last 7 days", "last 30 days"): Calculate StartDate = N days ago, EndDate = yesterday. Run date command separately first, then use the literal result in --start-date/--end-date. NEVER embed $(...) in CLI commands.
Execute queries based on user requirements:
Function 1: Query Total Daily Read/Write Usage
Query daily totals from StartDate to EndDate (default: last 30 days).
Output: Extract TotalUsageStatistics from the response.
Error Handling:
If the API returns an error indicating the project does not exist, recommend querying Function 2 first to get the complete list of queryable projects (deduplicate by project name).
Function 4: Query Daily Usage for a Specific Feature View
Query daily trends for a specific feature view from StartDate to EndDate (default: last 7 days).
Output: Extract the matching feature view from the FeatureViews array.
Error Handling:
If no matching feature view is found in the response, the feature view does not exist. Recommend querying Function 2 first to get the complete list of queryable feature views.
Step 4: Analyze Results and Calculate Costs
Based on user requirements:
Analyze usage trends from the query results
Calculate costs
Official Pricing (per 10,000 requests):
Region
Write Cost
Read Cost
Beijing, Shanghai, Hangzhou, Shenzhen
¥0.151
¥0.0755
Hong Kong, Singapore, Silicon Valley, Frankfurt, Jakarta
¥0.2651
¥0.1326
Present insights:
Identify high-traffic feature views
Highlight usage patterns (spikes, trends)
Provide cost breakdown and optimization suggestions if applicable
Offer next steps:
Ask if the user wants to drill down into specific projects or feature views
Offer to calculate costs if not already done
Suggest further analysis options
Success Verification
The skill execution is successful if:
✅ Valid PAI-FeatureStore instance is found
✅ FeatureDB datasource is identified or selected
✅ Usage statistics are successfully retrieved
✅ Results are analyzed and presented clearly
✅ Cost calculations are accurate (if requested)
✅ Response language matches user's question language
For detailed verification steps, see references/verification-method.md.
Cleanup
This skill performs read-only operations and does not create any resources. No cleanup is required.
Best Practices
Always confirm the RegionId with the user before executing any commands
Use pagination properly when listing datasources to ensure all results are retrieved
Provide context in analysis: Don't just show numbers—explain trends and patterns
Offer actionable insights: Suggest which feature views or projects to investigate further
Handle errors gracefully: If a project or feature view doesn't exist, guide the user to list all available options first
Calculate costs accurately: Use the correct pricing tier based on the region
Keep user engaged: After showing results, ask if they want to drill deeper or analyze specific aspects
Decline out-of-scope requests politely: If a request is not about querying usage statistics, politely decline and list supported features