UAT: TUI shell danger detection not wired to prompt — `ShellSafetyService` exists but visual warning never shown #5058

New issue

Closed

opened 2026-04-09 00:52:28 +00:00 by HAL9000 · 1 comment

HAL9000 commented

2026-04-09 00:52:28 +00:00

Owner

Bug Report

Feature Area: TUI — Shell Danger Detection / Safety Behaviors

Severity: Medium (safety warning not displayed)

What Was Tested

Shell danger detection as defined in the specification (§ Shell Danger Detection).

Expected Behavior (from spec)

When shell mode is active (!/$ prefix), the prompt performs heuristic analysis of the command to detect potentially destructive operations. Dangerous commands are highlighted with $error styling and a warning indicator appears below the prompt:

⚠ Potentially destructive command detected

The warning is advisory only — it never prevents command execution. Danger detection is controlled by shell.warn_dangerous setting (default: true).

Risk levels:

Pattern	Risk Level
`rm -rf` / `rm -r`	High
`chmod 777`	Medium
`> /dev/sda` / `dd if=`	High
Fork bomb patterns	High
`mkfs` / `fdisk` / `parted`	High
`kill -9` / `killall`	Medium
`sudo` / `su`	Low

Actual Behavior (from code analysis)

The ShellSafetyService and DangerousPatternDetector are fully implemented in src/cleveragents/tui/shell_safety/ with comprehensive pattern matching. However, they are never used by the TUI prompt.

The InputModeRouter in src/cleveragents/tui/input/modes.py uses a separate, simpler looks_dangerous() function from shell_exec.py (lines 21-31) that only checks 5 hardcoded patterns and blocks the command rather than showing a warning.

The ShellSafetyService is never instantiated or called anywhere in the TUI app code. No visual warning widget exists in the prompt area.

Steps to Reproduce

Run agents tui
Type !chmod 777 /var/www and press Enter
Observe: command executes without any visual warning
Type !sudo rm /etc/passwd and press Enter
Observe: command executes without any visual warning (only blocked if it matches the 5 hardcoded patterns in looks_dangerous())

Code Location

src/cleveragents/tui/shell_safety/safety_service.py — fully implemented but unused
src/cleveragents/tui/input/shell_exec.py lines 21-31 (looks_dangerous() — simple blocking, no visual warning)
src/cleveragents/tui/app.py — ShellSafetyService never imported or used

Impact

The spec-required visual warning for dangerous shell commands is never shown. Users receive no advisory feedback before executing potentially destructive commands. The ShellSafetyService implementation is dead code.

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

## Bug Report **Feature Area:** TUI — Shell Danger Detection / Safety Behaviors **Severity:** Medium (safety warning not displayed) ### What Was Tested Shell danger detection as defined in the specification (§ Shell Danger Detection). ### Expected Behavior (from spec) When shell mode is active (`!`/`$` prefix), the prompt performs heuristic analysis of the command to detect potentially destructive operations. Dangerous commands are highlighted with `$error` styling and a warning indicator appears below the prompt: > `⚠ Potentially destructive command detected` The warning is advisory only — it never prevents command execution. Danger detection is controlled by `shell.warn_dangerous` setting (default: `true`). Risk levels: | Pattern | Risk Level | |---------|-----------| | `rm -rf` / `rm -r` | High | | `chmod 777` | Medium | | `> /dev/sda` / `dd if=` | High | | Fork bomb patterns | High | | `mkfs` / `fdisk` / `parted` | High | | `kill -9` / `killall` | Medium | | `sudo` / `su` | Low | ### Actual Behavior (from code analysis) The `ShellSafetyService` and `DangerousPatternDetector` are fully implemented in `src/cleveragents/tui/shell_safety/` with comprehensive pattern matching. However, they are **never used by the TUI prompt**. The `InputModeRouter` in `src/cleveragents/tui/input/modes.py` uses a separate, simpler `looks_dangerous()` function from `shell_exec.py` (lines 21-31) that only checks 5 hardcoded patterns and **blocks** the command rather than showing a warning. The `ShellSafetyService` is never instantiated or called anywhere in the TUI app code. No visual warning widget exists in the prompt area. ### Steps to Reproduce 1. Run `agents tui` 2. Type `!chmod 777 /var/www` and press Enter 3. Observe: command executes without any visual warning 4. Type `!sudo rm /etc/passwd` and press Enter 5. Observe: command executes without any visual warning (only blocked if it matches the 5 hardcoded patterns in `looks_dangerous()`) ### Code Location - `src/cleveragents/tui/shell_safety/safety_service.py` — fully implemented but unused - `src/cleveragents/tui/input/shell_exec.py` lines 21-31 (`looks_dangerous()` — simple blocking, no visual warning) - `src/cleveragents/tui/app.py` — `ShellSafetyService` never imported or used ### Impact The spec-required visual warning for dangerous shell commands is never shown. Users receive no advisory feedback before executing potentially destructive commands. The `ShellSafetyService` implementation is dead code. --- **Automated by CleverAgents Bot** Supervisor: UAT Testing | Agent: uat-tester

HAL9000 commented

2026-04-09 00:56:32 +00:00

Author

Owner

Closing as duplicate of #4736 — both issues report the same problem: TUI shell danger detection not wired to ShellSafetyService.

Automated by CleverAgents Bot
Supervisor: Backlog Grooming | Agent: backlog-groomer

Closing as duplicate of #4736 — both issues report the same problem: TUI shell danger detection not wired to ShellSafetyService. --- **Automated by CleverAgents Bot** Supervisor: Backlog Grooming | Agent: backlog-groomer

HAL9000 closed this issue

2026-04-09 00:56:32 +00:00

HAL9000 referenced this issue

2026-04-09 00:58:39 +00:00

[AUTO-GROOMER] Backlog Grooming Report (Cycle 5) #5088

HAL9000 referenced this issue

2026-04-13 07:10:18 +00:00

docs(spec): clarify resource event types — add RESOURCE_CREATED, RESOURCE_DELETED, RESOURCE_MOVED #8038

HAL9000 referenced this issue

2026-04-13 18:57:29 +00:00

docs(spec): clarify resource event types — add RESOURCE_CREATED, RESOURCE_DELETED, RESOURCE_MOVED #8038