cleveragents/cleveragents-core

Fork 3

UAT: `agents session tell` is stubbed — echoes acknowledgement instead of invoking the bound actor #2583

New issue

Open

opened 2026-04-03 19:00:09 +00:00 by freemo · 1 comment

freemo commented

2026-04-03 19:00:09 +00:00

Owner

Metadata

Branch: fix/cli-session-tell-actor-invocation
Commit Message: fix(cli): implement real actor invocation in agents session tell
Milestone: v3.6.0
Parent Epic: #945

Background and Context

During UAT testing, a code analysis of cleveragents.cli.commands.session (the tell command handler) revealed that the agents session tell <PROMPT> --session <SESSION_ID> command is implemented as a stub. Instead of invoking the session's bound actor to generate a real AI response, it simply echoes a hardcoded acknowledgement string back to the user. The docstring in the implementation explicitly admits this: "For M3, the actor execution is stubbed — the assistant echoes an acknowledgement."

This means the primary purpose of the agents session tell command — enabling users to have conversations with AI actors via the CLI — is entirely non-functional.

Current Behavior

The implementation generates a hardcoded stub response instead of invoking the actor:

# Stub actor execution: generate simple assistant response
assistant_content = (
    f"Acknowledged: {prompt[:100]}"
    if not actor
    else f"[{actor}] Acknowledged: {prompt[:100]}"
)

Steps to reproduce:

Create a session: agents session create --actor openai/gpt-4
Send a message: agents session tell --session <ID> "What is 2+2?"
Observe: response is "Acknowledged: What is 2+2?" instead of an actual AI response

Code location: cleveragents.cli.commands.session — tell command handler, stub block at the actor execution step (lines 692, 709–719 at time of discovery).

Expected Behavior

Per the specification, agents session tell <PROMPT> --session <SESSION_ID> should:

Append the user message to the session
Invoke the session's bound actor (or the actor specified via --actor) to generate a real response
Append the actor's response to the session
Display the response (or stream it if --stream is used)

Acceptance Criteria

agents session tell invokes the session's bound actor and returns a real, non-stubbed response
The --actor override flag correctly substitutes a different actor for the invocation
The --stream flag streams the response in real-time
Actor-not-found and invocation errors are handled gracefully with clear, actionable error messages
The session history correctly records both the user message and the actor's response

Supporting Information

Code location at time of discovery: cleveragents.cli.commands.session — tell command handler, stub block (lines 692, 709–719)
Suggested fix: Replace the stub with actual actor invocation using the A2A facade or ReactiveCleverAgentsApp:
- Resolve the actor from the session's actor_name (or --actor override)
- Invoke the actor with the prompt
- Append the real response to the session
- Support streaming via --stream
Impact: The agents session tell command is non-functional for its primary purpose. Users cannot use the CLI to have conversations with actors via sessions.

Subtasks

Remove stub implementation from session tell command
Implement real actor invocation via A2A facade or ReactiveCleverAgentsApp
Support --stream flag for real-time response streaming
Handle actor-not-found and invocation errors gracefully
Add Behave unit tests with mocked actor
Add Robot Framework integration tests
Update PR and close this issue

Definition of Done

agents session tell invokes the bound actor and returns a real response
--stream flag streams the response in real-time
Error cases (no actor, actor unavailable) are handled with clear messages
All subtasks above are completed and checked off
A Git commit is created where the first line of the commit message matches the Commit Message in Metadata exactly, followed by a blank line, then additional lines providing relevant details about the implementation
The commit is pushed to the remote on the branch matching the Branch in Metadata exactly
The commit is submitted as a pull request to master, reviewed, and merged before this issue is marked done
All nox sessions pass
Coverage >= 97%

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

## Metadata - **Branch**: `fix/cli-session-tell-actor-invocation` - **Commit Message**: `fix(cli): implement real actor invocation in agents session tell` - **Milestone**: v3.6.0 - **Parent Epic**: #945 ## Background and Context During UAT testing, a code analysis of `cleveragents.cli.commands.session` (the `tell` command handler) revealed that the `agents session tell <PROMPT> --session <SESSION_ID>` command is implemented as a stub. Instead of invoking the session's bound actor to generate a real AI response, it simply echoes a hardcoded acknowledgement string back to the user. The docstring in the implementation explicitly admits this: *"For M3, the actor execution is stubbed — the assistant echoes an acknowledgement."* This means the primary purpose of the `agents session tell` command — enabling users to have conversations with AI actors via the CLI — is entirely non-functional. ## Current Behavior The implementation generates a hardcoded stub response instead of invoking the actor: ```python # Stub actor execution: generate simple assistant response assistant_content = ( f"Acknowledged: {prompt[:100]}" if not actor else f"[{actor}] Acknowledged: {prompt[:100]}" ) ``` **Steps to reproduce:** 1. Create a session: `agents session create --actor openai/gpt-4` 2. Send a message: `agents session tell --session <ID> "What is 2+2?"` 3. Observe: response is `"Acknowledged: What is 2+2?"` instead of an actual AI response **Code location:** `cleveragents.cli.commands.session` — `tell` command handler, stub block at the actor execution step (lines 692, 709–719 at time of discovery). ## Expected Behavior Per the specification, `agents session tell <PROMPT> --session <SESSION_ID>` should: 1. Append the user message to the session 2. Invoke the session's bound actor (or the actor specified via `--actor`) to generate a real response 3. Append the actor's response to the session 4. Display the response (or stream it if `--stream` is used) ## Acceptance Criteria - `agents session tell` invokes the session's bound actor and returns a real, non-stubbed response - The `--actor` override flag correctly substitutes a different actor for the invocation - The `--stream` flag streams the response in real-time - Actor-not-found and invocation errors are handled gracefully with clear, actionable error messages - The session history correctly records both the user message and the actor's response ## Supporting Information - **Code location at time of discovery**: `cleveragents.cli.commands.session` — `tell` command handler, stub block (lines 692, 709–719) - **Suggested fix**: Replace the stub with actual actor invocation using the A2A facade or `ReactiveCleverAgentsApp`: - Resolve the actor from the session's `actor_name` (or `--actor` override) - Invoke the actor with the prompt - Append the real response to the session - Support streaming via `--stream` - **Impact**: The `agents session tell` command is non-functional for its primary purpose. Users cannot use the CLI to have conversations with actors via sessions. ## Subtasks - [ ] Remove stub implementation from `session tell` command - [ ] Implement real actor invocation via A2A facade or ReactiveCleverAgentsApp - [ ] Support `--stream` flag for real-time response streaming - [ ] Handle actor-not-found and invocation errors gracefully - [ ] Add Behave unit tests with mocked actor - [ ] Add Robot Framework integration tests - [ ] Update PR and close this issue ## Definition of Done - [ ] `agents session tell` invokes the bound actor and returns a real response - [ ] `--stream` flag streams the response in real-time - [ ] Error cases (no actor, actor unavailable) are handled with clear messages - [ ] All subtasks above are completed and checked off - [ ] A Git commit is created where the **first line** of the commit message matches the Commit Message in Metadata exactly, followed by a blank line, then additional lines providing relevant details about the implementation - [ ] The commit is pushed to the remote on the branch matching the **Branch** in Metadata exactly - [ ] The commit is submitted as a **pull request** to `master`, reviewed, and **merged** before this issue is marked done - [ ] All nox sessions pass - [ ] Coverage >= 97% --- **Automated by CleverAgents Bot** Supervisor: UAT Testing | Agent: ca-uat-tester

freemo added this to the v3.6.0 milestone

2026-04-03 19:01:02 +00:00

freemo added a new dependency

2026-04-03 19:03:42 +00:00

#945 Legendary: Specification Alignment — Close All Gaps Between Implementation and Specification

freemo referenced this issue

2026-04-03 19:03:52 +00:00

[UAT-SUPERVISOR] Progress Report — Cycle 1591 — v3.7.0 #2593

freemo added a new dependency

2026-04-03 19:08:51 +00:00

#2597 fix(ci): restore all CI quality gates to passing on master — no suppression, no bypasses

freemo referenced this issue

2026-04-03 19:14:03 +00:00

[UAT-SUPERVISOR] Progress Report — Cycle 1601 — v3.7.0 #2603

freemo referenced this issue

2026-04-03 19:24:12 +00:00

[UAT-SUPERVISOR] Progress Report — Cycle 1611 — v3.7.0 #2605

freemo commented

2026-04-03 19:34:04 +00:00

Author

Owner

Issue triaged by project owner:

State: Verified
MoSCoW: Should Have

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

Issue triaged by project owner: - **State**: Verified - **MoSCoW**: Should Have --- **Automated by CleverAgents Bot** Supervisor: Project Owner | Agent: ca-project-owner

freemo referenced this issue

2026-04-03 19:34:18 +00:00

[UAT-SUPERVISOR] Progress Report — Cycle 1621 — v3.7.0 #2607

freemo referenced this issue

2026-04-03 19:44:25 +00:00

[UAT-SUPERVISOR] Progress Report — Cycle 1631 — v3.7.0 #2613

freemo referenced this issue

2026-04-03 19:54:34 +00:00

[UAT-SUPERVISOR] Progress Report — Cycle 1641 — v3.7.0 #2619

freemo referenced this issue

2026-04-03 20:04:40 +00:00

[UAT-SUPERVISOR] Progress Report — Cycle 1651 — v3.7.0 #2620