release: cut v0.7.4 with hook blocking fix

Remove type: "prompt" UserPromptSubmit hook that caused a small-model
gate to intermittently block legitimate user prompts with "Operation
stopped by hook" errors. The complexity-check command hook already
provides RLM activation guidance, making the prompt hook redundant.

Also includes all ty and ruff static analysis fixes from prior commit.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: rand

.claude-plugin/marketplace.json

@@ -10,7 +10,7 @@
       "name": "rlm-claude-code",
       "source": "./",
       "description": "Recursive Language Model integration for Claude Code - Go binary hooks (~5ms startup), cross-plugin event coordination (DP/RLM), version-aware config migration, and intelligent multi-provider routing",
-      "version": "0.7.3"
+      "version": "0.7.4"
     }
   ]
 }

.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "rlm-claude-code",
   "description": "Recursive Language Model integration for Claude Code - intelligent multi-provider routing and unbounded context handling",
-  "version": "0.7.3",
+  "version": "0.7.4",
   "author": {
     "name": "Rand",
     "url": "https://github.com/rand"

CHANGELOG.md

@@ -1,5 +1,14 @@
 # Changelog
 
+## [0.7.4] - 2026-02-20
+
+### Fixed
+- Remove `type: "prompt"` hook from UserPromptSubmit that caused small-model gate to block legitimate user prompts with "Operation stopped by hook" errors
+- Resolve all ty type checker diagnostics (4) and ruff linter errors (19) for clean static analysis
+
+### Changed
+- UserPromptSubmit now uses only the `complexity-check` command hook (no prompt-type gate)
+
 ## [0.7.3] - 2026-02-20
 
 ### Added

hooks/hooks.json

@@ -25,16 +25,6 @@
             "description": "Classify prompt complexity for RLM activation"
           }
         ]
-      },
-      {
-        "matcher": ".*",
-        "hooks": [
-          {
-            "type": "prompt",
-            "prompt": "You have access to RLM (Recursive Language Model) for complex tasks. If the user's request involves multi-file analysis, debugging complex issues, architectural decisions, or exhaustive searches, consider activating RLM with `/rlm mode balanced` or `/rlm mode thorough`. For simple tasks, RLM micro mode runs automatically. Do not mention RLM unless the task complexity warrants it.",
-            "description": "RLM activation guidance based on task complexity"
-          }
-        ]
       }
     ],
     "PreToolUse": [],

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "rlm-claude-code"
-version = "0.7.3"
+version = "0.7.4"
 description = "Recursive Language Model integration for Claude Code"
 readme = "README.md"
 requires-python = ">=3.12"

src/__init__.py

@@ -11,7 +11,7 @@
 - User-configurable preferences and budgets
 """
 
-__version__ = "0.7.3"
+__version__ = "0.7.4"
 
 # Async execution (SPEC-08.01-08.06) - from orchestrator package
 # Core orchestration