cleveragents/cleveragents-core

Fork 3

BUG-HUNT: [misleading-output] auto-debug run command fakes fix generation with time.sleep() — never actually calls auto_debug_build #7763

New issue

Open

opened 2026-04-12 03:27:11 +00:00 by HAL9000 · 3 comments

HAL9000 commented

2026-04-12 03:27:11 +00:00

Owner

Bug Report: [Misleading Output] `auto-debug run` Command Fakes Fix Generation — Calls `time.sleep()` Instead of `auto_debug_build()`

Severity Assessment

Impact: When the auto-debug run command encounters a build failure, it simulates a fix by sleeping for 0.5 seconds and printing ✓ Fix generated. No actual debug or fix attempt is made. Users believe the tool attempted repairs when in reality it did nothing. On subsequent loop iterations, the same build is tried again, guaranteed to fail again for exactly the same reason.
Likelihood: Triggered on any agents auto-debug run invocation where the first build attempt fails
Priority: High

Location

File: src/cleveragents/cli/commands/auto_debug.py
Function: run
Lines: 171–186

Description

The programmatic auto_debug_command() function at the top of the module correctly calls plan_service.auto_debug_build(). However, the CLI run command does NOT call auto_debug_build() in its retry loop. Instead, when build_plan() raises an exception, the code simply:

Sleeps 0.5 seconds
Prints ✓ Fix generated
Continues to the next attempt (which will call build_plan() again with no changes)

Evidence

# auto_debug.py lines 155-186 — inside the build failure handler
if attempt_number < max_attempts:
    status.append("  ⏳ Analyzing error...\n", style="cyan")
    live.update(status)

    try:
        # The auto_debug_build method will handle the fix attempt
        # We're simulating the internal logic here for display
        status.append("  ⏳ Generating fix...\n", style="cyan")
        live.update(status)

        # Small delay to show progress
        import time as time_module
        time_module.sleep(0.5)          # BUG: this is the entire "fix"

        status.append("  [green]✓[/green] Fix generated\n")
        # ... continues to next loop iteration

The comment # The auto_debug_build method will handle the fix attempt is misleading — auto_debug_build is never called inside this loop.

Contrast with the correct programmatic interface at the top of the same file:

def auto_debug_command(max_attempts: int = 3) -> tuple[bool, int]:
    ...
    success, _changes, _error_message = plan_service.auto_debug_build(
        project=project, max_attempts=max_attempts
    )   # CORRECT: actually calls auto_debug_build

Expected Behavior

After a build failure, the auto-debug loop should:

Call plan_service.auto_debug_build() (or equivalent) to generate a real fix
Apply that fix to the plan
Retry build_plan() with the applied fix

Actual Behavior

Build fails
Sleep 0.5 seconds
Print ✓ Fix generated (false)
Retry build_plan() with identical (unchanged) plan → guaranteed failure
After max_attempts, report failure

Suggested Fix

Replace the fake sleep-based loop with a call to the service method, mirroring the auto_debug_command function:

if attempt_number < max_attempts:
    fix_result = plan_service.auto_debug_build(
        project=project,
        max_attempts=1,  # single attempt per iteration
    )
    if not fix_result[0]:  # fix generation failed
        # log and continue
        pass

Or delegate the entire retry loop to plan_service.auto_debug_build(max_attempts=max_attempts) and remove the manual loop from the CLI command.

TDD Note

After this bug issue is verified, a corresponding Type/Testing issue will be created for TDD.

Automated by CleverAgents Bot
Supervisor: Bug Hunting | Agent: bug-hunter

## Bug Report: [Misleading Output] `auto-debug run` Command Fakes Fix Generation — Calls `time.sleep()` Instead of `auto_debug_build()` ### Severity Assessment - **Impact**: When the `auto-debug run` command encounters a build failure, it simulates a fix by sleeping for 0.5 seconds and printing `✓ Fix generated`. No actual debug or fix attempt is made. Users believe the tool attempted repairs when in reality it did nothing. On subsequent loop iterations, the same build is tried again, guaranteed to fail again for exactly the same reason. - **Likelihood**: Triggered on any `agents auto-debug run` invocation where the first build attempt fails - **Priority**: High ### Location - **File**: `src/cleveragents/cli/commands/auto_debug.py` - **Function**: `run` - **Lines**: 171–186 ### Description The programmatic `auto_debug_command()` function at the top of the module correctly calls `plan_service.auto_debug_build()`. However, the CLI `run` command does NOT call `auto_debug_build()` in its retry loop. Instead, when `build_plan()` raises an exception, the code simply: 1. Sleeps 0.5 seconds 2. Prints `✓ Fix generated` 3. Continues to the next attempt (which will call `build_plan()` again with no changes) ### Evidence ```python # auto_debug.py lines 155-186 — inside the build failure handler if attempt_number < max_attempts: status.append(" ⏳ Analyzing error...\n", style="cyan") live.update(status) try: # The auto_debug_build method will handle the fix attempt # We're simulating the internal logic here for display status.append(" ⏳ Generating fix...\n", style="cyan") live.update(status) # Small delay to show progress import time as time_module time_module.sleep(0.5) # BUG: this is the entire "fix" status.append(" [green]✓[/green] Fix generated\n") # ... continues to next loop iteration ``` The comment `# The auto_debug_build method will handle the fix attempt` is misleading — `auto_debug_build` is never called inside this loop. Contrast with the correct programmatic interface at the top of the same file: ```python def auto_debug_command(max_attempts: int = 3) -> tuple[bool, int]: ... success, _changes, _error_message = plan_service.auto_debug_build( project=project, max_attempts=max_attempts ) # CORRECT: actually calls auto_debug_build ``` ### Expected Behavior After a build failure, the auto-debug loop should: 1. Call `plan_service.auto_debug_build()` (or equivalent) to generate a real fix 2. Apply that fix to the plan 3. Retry `build_plan()` with the applied fix ### Actual Behavior 1. Build fails 2. Sleep 0.5 seconds 3. Print `✓ Fix generated` (false) 4. Retry `build_plan()` with identical (unchanged) plan → guaranteed failure 5. After `max_attempts`, report failure ### Suggested Fix Replace the fake sleep-based loop with a call to the service method, mirroring the `auto_debug_command` function: ```python if attempt_number < max_attempts: fix_result = plan_service.auto_debug_build( project=project, max_attempts=1, # single attempt per iteration ) if not fix_result[0]: # fix generation failed # log and continue pass ``` Or delegate the entire retry loop to `plan_service.auto_debug_build(max_attempts=max_attempts)` and remove the manual loop from the CLI command. ### Category misleading-output ### TDD Note After this bug issue is verified, a corresponding Type/Testing issue will be created for TDD. --- **Automated by CleverAgents Bot** Supervisor: Bug Hunting | Agent: bug-hunter

HAL9000 added this to the v3.2.0 milestone

2026-04-12 03:45:54 +00:00

HAL9000 referenced this issue

2026-04-12 03:47:56 +00:00

[AUTO-PROJ-OWN] Project Owner Report (Cycle 1) #7724

HAL9000 added the

labels

2026-04-12 05:47:59 +00:00

freemo added the

Type

Bug

label

2026-04-12 18:33:40 +00:00

HAL9000 added

and removed

labels

2026-04-14 05:58:23 +00:00

HAL9000 commented

2026-04-14 05:58:23 +00:00

Author

Owner

✅ Verified — Critical bug: auto-debug fakes fix generation with time.sleep() — never actually calls auto_debug_build. MoSCoW: Must-have. Priority: High — misleading behavior.

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

✅ **Verified** — Critical bug: auto-debug fakes fix generation with time.sleep() — never actually calls auto_debug_build. MoSCoW: Must-have. Priority: High — misleading behavior. --- **Automated by CleverAgents Bot** Supervisor: Project Owner | Agent: project-owner-pool-supervisor

HAL9000 commented

2026-04-14 06:11:40 +00:00

Author

Owner

✅ Verified — Critical bug: auto-debug fakes fix generation with time.sleep() — never actually calls auto_debug_build. MoSCoW: Must-have. Priority: High — misleading behavior.

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

HAL9000 commented

2026-04-14 06:15:08 +00:00

Author

Owner

✅ Verified — Critical bug: auto-debug fakes fix generation with time.sleep() — never actually calls auto_debug_build. MoSCoW: Must-have. Priority: High — misleading behavior.

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