UAT: `session tell --stream` simulates streaming character-by-character instead of using A2A `message/stream` SSE #4687

New issue

Open

opened 2026-04-08 18:00:55 +00:00 by HAL9000 · 0 comments

HAL9000 commented

2026-04-08 18:00:55 +00:00

Owner

Summary

The --stream flag on agents session tell simulates streaming by iterating over characters of the stub response string. The spec requires real streaming via the A2A message/stream operation returning TaskStatusUpdateEvent / TaskArtifactUpdateEvent via SSE.

Expected Behavior (from `docs/specification.md` §agents session tell)

When --stream is used:

A "Session" panel is shown with ID, Actor, Mode: streaming
Response tokens stream in real time as the orchestrator works
A "Usage" panel shows Tokens, Duration, Tool Calls
Success line: ✓ OK Stream complete

The spec maps --stream to message/stream A2A operation:

message/stream | Client → Server | SessionWorkflow.tell() with streaming — returns TaskStatusUpdateEvent / TaskArtifactUpdateEvent via SSE

Actual Behavior

src/cleveragents/cli/commands/session.py — tell() function (lines 840–850):

if stream:
    # Simulate streaming by printing character by character
    for char in assistant_content:
        sys.stdout.write(char)
        sys.stdout.flush()
    sys.stdout.write("\n")

This:

Does NOT use A2A message/stream
Does NOT produce SSE events
Does NOT show the spec-required "Session" panel with Mode: streaming
Does NOT show the "Usage" panel
Simply iterates over characters of the stub "Acknowledged: ..." string

Code Location

src/cleveragents/cli/commands/session.py — tell() function, lines 840–850

Impact

Users cannot see real-time orchestrator progress
The --stream flag provides no functional benefit over non-streaming mode
The spec's primary streaming use case is completely unimplemented

Automated by CleverAgents Bot
Supervisor: UAT Testing | Agent: uat-tester

## Summary The `--stream` flag on `agents session tell` simulates streaming by iterating over characters of the stub response string. The spec requires real streaming via the A2A `message/stream` operation returning `TaskStatusUpdateEvent` / `TaskArtifactUpdateEvent` via SSE. ## Expected Behavior (from `docs/specification.md` §agents session tell) When `--stream` is used: 1. A "Session" panel is shown with ID, Actor, Mode: streaming 2. Response tokens stream in real time as the orchestrator works 3. A "Usage" panel shows Tokens, Duration, Tool Calls 4. Success line: `✓ OK Stream complete` The spec maps `--stream` to `message/stream` A2A operation: > `message/stream` | Client → Server | `SessionWorkflow.tell()` with streaming — returns `TaskStatusUpdateEvent` / `TaskArtifactUpdateEvent` via SSE ## Actual Behavior `src/cleveragents/cli/commands/session.py` — `tell()` function (lines 840–850): ```python if stream: # Simulate streaming by printing character by character for char in assistant_content: sys.stdout.write(char) sys.stdout.flush() sys.stdout.write("\n") ``` This: 1. Does NOT use A2A `message/stream` 2. Does NOT produce SSE events 3. Does NOT show the spec-required "Session" panel with Mode: streaming 4. Does NOT show the "Usage" panel 5. Simply iterates over characters of the stub "Acknowledged: ..." string ## Code Location - `src/cleveragents/cli/commands/session.py` — `tell()` function, lines 840–850 ## Impact - Users cannot see real-time orchestrator progress - The `--stream` flag provides no functional benefit over non-streaming mode - The spec's primary streaming use case is completely unimplemented --- **Automated by CleverAgents Bot** Supervisor: UAT Testing | Agent: uat-tester

Rows
Columns

UAT: session tell --stream simulates streaming character-by-character instead of using A2A message/stream SSE #4687