refactor(agent-system): replace test-verification hooks with /verify-tests skill

The TaskCompleted, implementer Stop, refactorer PostToolUse, and test-writer
Stop hooks each ran the project test suite automatically on agent/task
lifecycle events, serializing 30-120s of verification work per event.

Replace all four with a single on-demand /verify-tests skill that Claude
invokes explicitly. Keeps test-running behaviour available without forcing
it on every stop or completion, aligning with the same pattern used by /cq.

Author: AnExiledDev

container/.devcontainer/AGENTS.md

@@ -48,7 +48,7 @@ Declared in `settings.json` under `enabledPlugins`, auto-activated on start:
 
 ### Active
 
-- **agent-system** — 4 custom agents (architect, claude-guide, explorer, generalist) + built-in agent redirection
+- **agent-system** — 4 custom agents (architect, claude-guide, explorer, generalist) + built-in agent redirection + `/verify-tests` skill
 - **skill-engine** — 2 coding knowledge packs (`/team`, `/agent-browser`) + auto-suggestion
 - **auto-code-quality** — File tracking, syntax validation, `/cq` quality gate (format + lint + test on demand)
 - **session-context** — Git state injection, TODO harvesting, commit reminders

container/.devcontainer/CHANGELOG.md

@@ -53,6 +53,7 @@
 ### Agent System
 
 - **All agents and skills set to `effort: max`** — active agents (claude-guide, explorer, generalist), all 14 archived agents in `agent-system/agents/_archived/`, both active skill-engine skills (`team`, `agent-browser`), all 22 archived skills across `skill-engine` and `agent-system`, and all 10 skills in currently disabled plugins (`git-workflow`, `prompt-snippets`, `spec-workflow`, `ticket-workflow`) now declare `effort: max`. This ensures any plugin re-enable in the future yields max-effort runs without further edits.
+- **Replaced automatic test hooks with `/verify-tests` skill** — removed `TaskCompleted` hook (`task-completed-check.py`), implementer Stop hook (`verify-no-regression.py`), refactorer PostToolUse hook (`verify-no-regression.py`), and test-writer Stop hook (`verify-tests-pass.py`). Test verification is now on-demand via `/verify-tests`, which detects the project's test framework and runs the suite with structured reporting.
 
 ### Plugin Cleanup
 

container/.devcontainer/plugins/devs-marketplace/.claude-plugin/marketplace.json

@@ -52,7 +52,7 @@
 		},
 		{
 			"name": "agent-system",
-			"description": "4 custom agents with built-in agent redirection (15 archived for rewrite)",
+			"description": "4 custom agents with built-in agent redirection and /verify-tests skill (15 archived for rewrite)",
 			"version": "1.0.0",
 			"source": "./plugins/agent-system",
 			"category": "development",

container/.devcontainer/plugins/devs-marketplace/plugins/agent-system/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
 	"name": "agent-system",
-	"description": "4 custom agents with built-in agent redirection (15 archived for rewrite)",
+	"description": "4 custom agents with built-in agent redirection and /verify-tests skill (15 archived for rewrite)",
 	"author": {
 		"name": "AnExiledDev"
 	}

container/.devcontainer/plugins/devs-marketplace/plugins/agent-system/README.md

@@ -1,10 +1,10 @@
 # agent-system
 
-Claude Code plugin that provides 4 custom agents with automatic built-in agent redirection, working directory injection, read-only bash enforcement, and team quality gates. 15 additional agents are archived in `agents/_archived/` pending rewrite.
+Claude Code plugin that provides 4 custom agents with automatic built-in agent redirection, working directory injection, read-only bash enforcement, `/verify-tests` skill, and team quality gates. 15 additional agents are archived in `agents/_archived/` pending rewrite.
 
 ## What It Does
 
-Replaces Claude Code's built-in agents with enhanced custom agents that carry domain-specific instructions, safety hooks, and tailored tool configurations. Also provides team orchestration quality gates.
+Replaces Claude Code's built-in agents with enhanced custom agents that carry domain-specific instructions, safety hooks, and tailored tool configurations. Includes `/verify-tests` for on-demand test suite execution.
 
 ### Active Agents
 
@@ -32,12 +32,18 @@ The following agents are preserved in `agents/_archived/` for future rewrite:
 
 bash-exec, debug-logs, dependency-analyst, documenter, git-archaeologist, implementer, investigator, migrator, perf-profiler, refactorer, researcher, security-auditor, spec-writer, statusline-config, test-writer
 
-### Quality Gates
+### Skills
+
+| Skill | Purpose |
+|-------|---------|
+| `/debug` | Structured log investigation and diagnosis |
+| `/verify-tests` | Run project test suite, report results, fix failures |
+
+### Orchestration Hooks
 
 | Hook | Script | Purpose |
 |------|--------|---------|
 | TeammateIdle | `teammate-idle-check.py` | Prevents teammates from going idle with incomplete tasks |
-| TaskCompleted | `task-completed-check.py` | Runs test suite before allowing task completion |
 
 ## How It Works
 
@@ -55,14 +61,6 @@ Claude calls the Task tool (spawning a subagent)
   |
   +-> Subagent works...
   |
-  +-> TaskCompleted fires
-  |     |
-  |     +-> task-completed-check.py
-  |           |
-  |           +-> Detect test framework -> Run tests
-  |           +-> Tests pass? -> Allow completion
-  |           +-> Tests fail? -> Block, send feedback
-  |
   +-> TeammateIdle fires (team mode)
         |
         +-> teammate-idle-check.py
@@ -86,7 +84,6 @@ Read-only agents (explorer, architect) have their Bash access restricted by `gua
 |--------|--------|--------|
 | redirect-builtin-agents.py | Allow (or rewrite) | Block with error |
 | guard-readonly-bash.py | Allow command | Block write operation |
-| task-completed-check.py | Tests pass | Tests fail (block completion) |
 | teammate-idle-check.py | No incomplete tasks | Has incomplete tasks |
 
 ### Timeouts
@@ -95,7 +92,6 @@ Read-only agents (explorer, architect) have their Bash access restricted by `gua
 |------|---------|
 | Agent redirection (PreToolUse) | 5s |
 | Teammate idle check | 10s |
-| Task completed check | 60s |
 
 ## Installation
 
@@ -157,14 +153,13 @@ agent-system/
 +-- scripts/
 |   +-- guard-readonly-bash.py       # Read-only bash enforcement
 |   +-- redirect-builtin-agents.py   # Built-in agent redirection
-|   +-- task-completed-check.py      # Test suite quality gate
 |   +-- teammate-idle-check.py       # Incomplete task checker
-|   +-- verify-no-regression.py      # Post-edit regression tests (dormant — agents archived)
-|   +-- verify-tests-pass.py         # Test verification (dormant — agents archived)
 +-- skills/
 |   +-- _archived/
-|       +-- debug/
-|           +-- SKILL.md             # Log investigation skill (archived)
+|   |   +-- debug/
+|   |       +-- SKILL.md             # Log investigation skill (archived)
+|   +-- verify-tests/
+|       +-- SKILL.md                 # On-demand test suite runner
 +-- AGENT-REDIRECTION.md             # Redirection mechanism docs
 +-- REVIEW-RUBRIC.md                 # Agent/skill quality rubric
 +-- README.md                        # This file

container/.devcontainer/plugins/devs-marketplace/plugins/agent-system/hooks/hooks.json

@@ -1,5 +1,5 @@
 {
-	"description": "Agent redirection, subagent configuration, and team quality gate hooks",
+	"description": "Agent redirection and team orchestration hooks",
 	"hooks": {
 		"PreToolUse": [
 			{
@@ -23,17 +23,6 @@
 					}
 				]
 			}
-		],
-		"TaskCompleted": [
-			{
-				"hooks": [
-					{
-						"type": "command",
-						"command": "python3 ${CLAUDE_PLUGIN_ROOT}/scripts/task-completed-check.py",
-						"timeout": 60
-					}
-				]
-			}
 		]
 	}
 }