Files
IdeA/.ideai/tickets/156/issue.md

17 lines
963 B
Markdown

---
id: "8ba27857-773a-4303-a7b5-ae7d688b3442"
number: 156
title: "Foundation: unified streaming progress/event model across AI profiles"
status: "closed"
priority: "high"
sprint: null
links: []
agentRefs: [{"agentId":"fe887179-933f-47d4-960f-c3b06827f86c","role":"assigned"}]
attachments: []
createdBy: {"kind":"agent","agent_id":"a6c6ea12-bfc6-4bdc-8031-324102dfa34d"}
updatedBy: {"kind":"agent","agent_id":"1ff94b51-3e17-4a39-8543-7715d1ff0f80"}
createdAt: 1785965383888
updatedAt: 1786014830849
version: 7
---
Define and implement a provider-agnostic progress/event pipeline so IdeA can expose as much non-terminal activity as each AI profile makes available before Final. Scope includes a canonical event taxonomy for chat/delegation flows, mapping from profile-specific adapters (Codex, Claude, OpenAI-style structured, and future profiles), and transport/projection rules that preserve robustness when some profiles expose little or no intermediate output.