Query id_changes logs to track identity stitching events and reconstruct profile histories.
npx skills add https://github.com/treasure-data/td-skills --skill identity
Please help me install this Agent Skill. Skill: identity Source: https://github.com/treasure-data/td-skills/tree/main/realtime-skills/identity Command: npx skills add https://github.com/treasure-data/td-skills --skill identity
Query identity changes logs to understand profile creation, updates, and merging in real-time within a streaming real-time context.
Query the id_changes log for the last 24 hours with a sample query.
High-intent search queries and answers about installing and using this skill.
Query the id_changes log using a numeric parent segment ID to retrieve stitching events like profile_added, profile_updated_by_stitching, and profile_deleted_by_stitching. Access requires tdx-skill or Treasure Data mcp-server, an API key with privileges to the cdp_audience_<segment>_rt table, and time-bounded queries using TD_INTERVAL and TD_TIME_FORMAT.
Identity stitching logs record profile creation, merging, and deletion events. You can reconstruct profile histories, understand identity flows across updates, extract key-value attributes, and investigate segment-specific changes within configurable time ranges to track how profiles are created and consolidated.
You need a properly configured tdx-skill or Treasure Data mcp-server access, an API key with sufficient privileges to the parent-segment id_changes table formatted as cdp_audience_<segment>_rt, and the numeric parent segment ID. Time-bounded queries require support for TD_INTERVAL and TD_TIME_FORMAT.
Yes. Identity stitching logs support real-time workflows for segment-specific investigations. Query profile changes within specific time ranges to understand stitching behavior, track merging events, and analyze identity flows as they occur in your streaming context.
Use TD_INTERVAL and TD_TIME_FORMAT parameters to bound your queries by time range. These enable you to query the id_changes table for specific periods, such as the last 24 hours, to isolate profile events and reconstruct histories within your investigation window.
treasure-dataQuery Treasure Data activation logs to detect delivery errors.
DanhvbAnalyze CDP concepts for data unification, identity resolution, and audience activation.
huivrotikiAutomate Segment tracking, identifying, batching, grouping, page views, and aliasing via MCP.
yunqiqiliangOrchestrate Table Stream CDC pipelines with change_tracking and MERGE consumption.
