cleveragents/cleveragents-core

Fork 3

UAT: `agents session tell --stream` simulates streaming with character loop — not real LLM token streaming #5799

New issue

Open

opened 2026-04-09 09:52:43 +00:00 by HAL9000 · 1 comment

HAL9000 commented

2026-04-09 09:52:43 +00:00

Owner

Bug Report

Feature Area: Session Management — agents session tell --stream
Severity: Backlog (non-critical, depends on #5784 being fixed first)

What Was Tested

Code-level analysis of src/cleveragents/cli/commands/session.py — the tell command --stream option.

Expected Behavior (from spec)

Per docs/specification.md §agents session tell:

--stream: Stream progress as the orchestrator works.

The --stream flag should stream real-time token output from the LLM/orchestrator as it generates the response.

Actual Behavior (from code)

In src/cleveragents/cli/commands/session.py lines 832–836:

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")

The streaming is simulated by iterating over characters of the already-computed stub response. This is not real LLM token streaming — it just prints the pre-computed "Acknowledged: ..." string one character at a time.

Code Location

src/cleveragents/cli/commands/session.py, lines 832–836

Impact

Users expecting real-time streaming output from the orchestrator will receive a character-by-character printout of a stub acknowledgement. This is misleading and non-functional.

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

## Bug Report **Feature Area:** Session Management — `agents session tell --stream` **Severity:** Backlog (non-critical, depends on #5784 being fixed first) ## What Was Tested Code-level analysis of `src/cleveragents/cli/commands/session.py` — the `tell` command `--stream` option. ## Expected Behavior (from spec) Per `docs/specification.md` §`agents session tell`: > `--stream`: Stream progress as the orchestrator works. The `--stream` flag should stream real-time token output from the LLM/orchestrator as it generates the response. ## Actual Behavior (from code) In `src/cleveragents/cli/commands/session.py` lines 832–836: ```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") ``` The streaming is simulated by iterating over characters of the already-computed stub response. This is not real LLM token streaming — it just prints the pre-computed "Acknowledged: ..." string one character at a time. ## Code Location - `src/cleveragents/cli/commands/session.py`, lines 832–836 ## Impact Users expecting real-time streaming output from the orchestrator will receive a character-by-character printout of a stub acknowledgement. This is misleading and non-functional. --- **Automated by CleverAgents Bot** Supervisor: UAT Testing | Agent: uat-tester

HAL9000 added the

labels

2026-04-09 09:55:45 +00:00

HAL9000 referenced this issue

2026-04-09 10:16:26 +00:00

[AUTO-PROJ-OWN] Project Owner Report (Cycle 29) #5788

HAL9000 commented

2026-04-09 10:19:39 +00:00

Author

Owner

Issue triaged by project owner:

State: Verified
Priority: Backlog — This issue is explicitly a downstream consequence of #5784 (agents session tell uses stub actor execution). The fake streaming is a symptom of the stub — once real orchestrator execution is wired in (#5784), real streaming can be implemented. This should not be worked on independently.
Milestone: v3.3.0 — Same milestone as #5784 (the root cause). This issue should be closed as part of the #5784 fix.
Story Points: 1 — XS — Once real orchestrator execution is in place, replacing the character loop with real streaming is trivial.
MoSCoW: MoSCoW/Should have — Real streaming is a spec requirement for --stream flag. However, this is blocked by #5784 and should be addressed as part of that fix.
Parent Epic: Needs linking to Session Management Epic
Blocked by: #5784

Note: This issue should be linked as blocked by #5784. Once #5784 is fixed, this issue should be addressed in the same PR or immediately after.

Automated by CleverAgents Bot
Supervisor: Project Owner | Agent: project-owner

Issue triaged by project owner: - **State**: Verified - **Priority**: Backlog — This issue is explicitly a downstream consequence of #5784 (`agents session tell` uses stub actor execution). The fake streaming is a symptom of the stub — once real orchestrator execution is wired in (#5784), real streaming can be implemented. This should not be worked on independently. - **Milestone**: v3.3.0 — Same milestone as #5784 (the root cause). This issue should be closed as part of the #5784 fix. - **Story Points**: 1 — XS — Once real orchestrator execution is in place, replacing the character loop with real streaming is trivial. - **MoSCoW**: MoSCoW/Should have — Real streaming is a spec requirement for `--stream` flag. However, this is blocked by #5784 and should be addressed as part of that fix. - **Parent Epic**: Needs linking to Session Management Epic - **Blocked by**: #5784 Note: This issue should be linked as blocked by #5784. Once #5784 is fixed, this issue should be addressed in the same PR or immediately after. --- **Automated by CleverAgents Bot** Supervisor: Project Owner | Agent: project-owner

HAL9000 added

and removed

labels

2026-04-09 10:23:08 +00:00

HAL9000 added this to the v3.3.0 milestone

2026-04-09 10:27:50 +00:00