Skip to content

Add unit_primitives field for locally define prompts

Jaydip Pansuriyarequested to merge
578228-fix-flow-v2-validation-file into master

What does this MR do and why?

I discovered that in the current AI catalog flow's flow_v2 version (app/validators/json_schemas/ai_catalog/flow_v2.json), unit_primitives is not included for locally defined prompts. Instead, it is part of the model object. This MR adds unit_primitives as an optional field under the prompts object and removes it from the model component. For reference, see the flow register v1 example.

version: "v1"
environment: ambient

components:
    - name: "code_analyzer"
      type: AgentComponent
      prompt_id: "my_local_prompt"
      # prompt_version omitted - uses local prompt
      ui_log_events: [ "on_agent_final_answer" ]
      inputs: ["context:goal"]
      toolset: ["read_file", "list_dir"]

prompts:
    - prompt_id: "my_local_prompt"
      name: "Code Analysis Assistant"
      model:
        params:
            model_class_provider: anthropic
            model: claude-sonnet-4-20250514
            max_tokens: 32_768
      unit_primitives: []
      prompt_template:
          system: "You are a helpful assistant specialized in code analysis"
          user: "{{goal}}"
          placeholder: history  # Optional: include conversation history
      params:
          timeout: 180

routers:
    - from: "code_analyzer"
      to: "end"

flow:
    entry_point: "code_analyzer"

Note: Currently, there are no flow records in production, as this feature is behind a feature flag set to false by default.

 Index Scan using index_ai_catalog_items_on_item_type on public.ai_catalog_items  (cost=0.15..3.17 rows=1 width=273) (actual time=0.022..0.023 rows=0 loops=1)
   Index Cond: (ai_catalog_items.item_type = 2)
   Buffers: shared hit=4
   I/O Timings: read=0.000 write=0.000
Settings: seq_page_cost = '4', work_mem = '100MB', random_page_cost = '1.5', jit = 'off', effective_cache_size = '472585MB'

References

Screenshots or screen recordings

Before After

How to set up and validate locally

MR acceptance checklist

Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.

Related to #578228

Merge request reports