gobby 0.2.5__py3-none-any.whl → 0.2.6__py3-none-any.whl

gobby/mcp_proxy/tools/skills/__init__.py

@@ -0,0 +1,631 @@
+"""
+Internal MCP tools for Skill management.
+
+Exposes functionality for:
+- list_skills(): List all skills with lightweight metadata
+- get_skill(): Get skill by ID or name with full content
+- search_skills(): Search skills by query with relevance ranking
+- update_skill(): Update an existing skill by refreshing from source
+- install_skill(): Install skill from local path, GitHub URL, or ZIP archive
+- remove_skill(): Remove a skill by name or ID
+
+These tools use LocalSkillManager for storage and SkillSearch for search.
+"""
+
+from __future__ import annotations
+
+import re
+from collections.abc import Callable
+from pathlib import Path
+from typing import TYPE_CHECKING, Any
+
+from gobby.mcp_proxy.tools.internal import InternalToolRegistry
+from gobby.skills.loader import SkillLoader, SkillLoadError
+from gobby.skills.search import SearchFilters, SkillSearch
+from gobby.skills.updater import SkillUpdater
+from gobby.storage.skills import LocalSkillManager
+
+if TYPE_CHECKING:
+    from gobby.storage.database import DatabaseProtocol
+
+__all__ = ["create_skills_registry", "SkillsToolRegistry"]
+
+
+class SkillsToolRegistry(InternalToolRegistry):
+    """Registry for skill management tools with test-friendly get_tool method."""
+
+    search: SkillSearch  # Assigned dynamically in create_skills_registry
+
+    def get_tool(self, name: str) -> Callable[..., Any] | None:
+        """Get a tool function by name (for testing)."""
+        tool = self._tools.get(name)
+        return tool.func if tool else None
+
+
+def create_skills_registry(
+    db: DatabaseProtocol,
+    project_id: str | None = None,
+) -> SkillsToolRegistry:
+    """
+    Create a skills management tool registry.
+
+    Args:
+        db: Database connection for storage
+        project_id: Optional default project scope for skill operations
+
+    Returns:
+        SkillsToolRegistry with skill management tools registered
+    """
+    registry = SkillsToolRegistry(
+        name="gobby-skills",
+        description="Skill management - list_skills, get_skill, search_skills, install_skill, update_skill, remove_skill",
+    )
+
+    # Initialize storage
+    storage = LocalSkillManager(db)
+
+    # --- list_skills tool ---
+
+    @registry.tool(
+        name="list_skills",
+        description="List all skills with lightweight metadata. Supports filtering by category and enabled status.",
+    )
+    async def list_skills(
+        category: str | None = None,
+        enabled: bool | None = None,
+        limit: int = 50,
+    ) -> dict[str, Any]:
+        """
+        List skills with lightweight metadata.
+
+        Returns ~100 tokens per skill: name, description, category, tags, enabled.
+        Does NOT include content, allowed_tools, or compatibility.
+
+        Args:
+            category: Optional category filter
+            enabled: Optional enabled status filter (True/False/None for all)
+            limit: Maximum skills to return (default 50)
+
+        Returns:
+            Dict with success status and list of skill metadata
+        """
+        try:
+            skills = storage.list_skills(
+                project_id=project_id,
+                category=category,
+                enabled=enabled,
+                limit=limit,
+                include_global=True,
+            )
+
+            # Extract lightweight metadata only
+            skill_list = []
+            for skill in skills:
+                # Get category and tags from metadata
+                category_value = None
+                tags = []
+                if skill.metadata and isinstance(skill.metadata, dict):
+                    skillport = skill.metadata.get("skillport", {})
+                    if isinstance(skillport, dict):
+                        category_value = skillport.get("category")
+                        tags = skillport.get("tags", [])
+
+                skill_list.append(
+                    {
+                        "id": skill.id,
+                        "name": skill.name,
+                        "description": skill.description,
+                        "category": category_value,
+                        "tags": tags,
+                        "enabled": skill.enabled,
+                    }
+                )
+
+            return {
+                "success": True,
+                "count": len(skill_list),
+                "skills": skill_list,
+            }
+        except Exception as e:
+            return {
+                "success": False,
+                "error": str(e),
+            }
+
+    # --- get_skill tool ---
+
+    @registry.tool(
+        name="get_skill",
+        description="Get full skill content by name or ID. Returns complete skill including content, allowed_tools, etc.",
+    )
+    async def get_skill(
+        name: str | None = None,
+        skill_id: str | None = None,
+    ) -> dict[str, Any]:
+        """
+        Get a skill by name or ID with full content.
+
+        Returns all skill fields including content, allowed_tools, compatibility.
+        Use this after list_skills to get the full skill when needed.
+
+        Args:
+            name: Skill name (used if skill_id not provided)
+            skill_id: Skill ID (takes precedence over name)
+
+        Returns:
+            Dict with success status and full skill data
+        """
+        try:
+            # Validate input
+            if not skill_id and not name:
+                return {
+                    "success": False,
+                    "error": "Either name or skill_id is required",
+                }
+
+            # Get skill by ID or name
+            skill = None
+            if skill_id:
+                try:
+                    skill = storage.get_skill(skill_id)
+                except ValueError:
+                    pass
+
+            if skill is None and name:
+                skill = storage.get_by_name(name, project_id=project_id)
+
+            if skill is None:
+                return {
+                    "success": False,
+                    "error": f"Skill not found: {skill_id or name}",
+                }
+
+            # Return full skill data
+            return {
+                "success": True,
+                "skill": {
+                    "id": skill.id,
+                    "name": skill.name,
+                    "description": skill.description,
+                    "content": skill.content,
+                    "version": skill.version,
+                    "license": skill.license,
+                    "compatibility": skill.compatibility,
+                    "allowed_tools": skill.allowed_tools,
+                    "metadata": skill.metadata,
+                    "enabled": skill.enabled,
+                    "source_path": skill.source_path,
+                    "source_type": skill.source_type,
+                    "source_ref": skill.source_ref,
+                },
+            }
+        except Exception as e:
+            return {
+                "success": False,
+                "error": str(e),
+            }
+
+    # --- search_skills tool ---
+
+    # Initialize search and index skills
+    search = SkillSearch()
+    # Expose search instance on registry for testing/manual indexing
+    registry.search = search
+
+    def _index_skills() -> None:
+        """Index all skills for search."""
+        skills = storage.list_skills(
+            project_id=project_id,
+            limit=10000,
+            include_global=True,
+        )
+        search.index_skills(skills)
+
+    # Index on registry creation
+    _index_skills()
+
+    @registry.tool(
+        name="search_skills",
+        description="Search for skills by query. Returns ranked results with relevance scores. Supports filtering by category and tags.",
+    )
+    async def search_skills(
+        query: str,
+        category: str | None = None,
+        tags_any: list[str] | None = None,
+        tags_all: list[str] | None = None,
+        top_k: int = 10,
+    ) -> dict[str, Any]:
+        """
+        Search for skills by natural language query.
+
+        Returns ranked results with relevance scores.
+
+        Args:
+            query: Search query (required, non-empty)
+            category: Optional category filter
+            tags_any: Optional tags filter - match any of these tags
+            tags_all: Optional tags filter - match all of these tags
+            top_k: Maximum results to return (default 10)
+
+        Returns:
+            Dict with success status and ranked search results
+        """
+        try:
+            # Validate query
+            if not query or not query.strip():
+                return {
+                    "success": False,
+                    "error": "Query is required and cannot be empty",
+                }
+
+            # Build filters
+            filters = None
+            if category or tags_any or tags_all:
+                filters = SearchFilters(
+                    category=category,
+                    tags_any=tags_any,
+                    tags_all=tags_all,
+                )
+
+            # Perform search
+            results = await search.search_async(query=query, top_k=top_k, filters=filters)
+
+            # Format results with skill metadata
+            result_list = []
+            for r in results:
+                # Look up skill to get description, category, tags
+                skill = None
+                try:
+                    skill = storage.get_skill(r.skill_id)
+                except ValueError:
+                    pass
+
+                # Get category and tags from metadata
+                category_value = None
+                tags = []
+                if skill and skill.metadata and isinstance(skill.metadata, dict):
+                    skillport = skill.metadata.get("skillport", {})
+                    if isinstance(skillport, dict):
+                        category_value = skillport.get("category")
+                        tags = skillport.get("tags", [])
+
+                result_list.append(
+                    {
+                        "skill_id": r.skill_id,
+                        "skill_name": r.skill_name,
+                        "description": skill.description if skill else None,
+                        "category": category_value,
+                        "tags": tags,
+                        "score": r.similarity,
+                    }
+                )
+
+            return {
+                "success": True,
+                "count": len(result_list),
+                "results": result_list,
+            }
+        except Exception as e:
+            return {
+                "success": False,
+                "error": str(e),
+            }
+
+    # --- remove_skill tool ---
+
+    @registry.tool(
+        name="remove_skill",
+        description="Remove a skill by name or ID. Returns success status and removed skill name.",
+    )
+    async def remove_skill(
+        name: str | None = None,
+        skill_id: str | None = None,
+    ) -> dict[str, Any]:
+        """
+        Remove a skill from the database.
+
+        Args:
+            name: Skill name (used if skill_id not provided)
+            skill_id: Skill ID (takes precedence over name)
+
+        Returns:
+            Dict with success status and removed skill info
+        """
+        try:
+            # Validate input
+            if not skill_id and not name:
+                return {
+                    "success": False,
+                    "error": "Either name or skill_id is required",
+                }
+
+            # Find the skill first to get its name
+            skill = None
+            if skill_id:
+                try:
+                    skill = storage.get_skill(skill_id)
+                except ValueError:
+                    pass
+
+            if skill is None and name:
+                skill = storage.get_by_name(name, project_id=project_id)
+
+            if skill is None:
+                return {
+                    "success": False,
+                    "error": f"Skill not found: {skill_id or name}",
+                }
+
+            # Store the name before deletion
+            skill_name = skill.name
+
+            # Delete the skill
+            storage.delete_skill(skill.id)
+
+            # Re-index skills after deletion
+            skills = storage.list_skills(
+                project_id=project_id,
+                limit=10000,
+                include_global=True,
+            )
+            await search.index_skills_async(skills)
+
+            return {
+                "success": True,
+                "removed": True,
+                "skill_name": skill_name,
+            }
+        except Exception as e:
+            return {
+                "success": False,
+                "error": str(e),
+            }
+
+    # --- update_skill tool ---
+
+    # Initialize updater
+    updater = SkillUpdater(storage)
+
+    @registry.tool(
+        name="update_skill",
+        description="Update a skill by refreshing from its source. Returns whether the skill was updated.",
+    )
+    async def update_skill(
+        name: str | None = None,
+        skill_id: str | None = None,
+    ) -> dict[str, Any]:
+        """
+        Update a skill by refreshing from its source path.
+
+        Args:
+            name: Skill name (used if skill_id not provided)
+            skill_id: Skill ID (takes precedence over name)
+
+        Returns:
+            Dict with success status and update info
+        """
+        try:
+            # Validate input
+            if not skill_id and not name:
+                return {
+                    "success": False,
+                    "error": "Either name or skill_id is required",
+                }
+
+            # Find the skill first
+            skill = None
+            if skill_id:
+                try:
+                    skill = storage.get_skill(skill_id)
+                except ValueError:
+                    pass
+
+            if skill is None and name:
+                skill = storage.get_by_name(name, project_id=project_id)
+
+            if skill is None:
+                return {
+                    "success": False,
+                    "error": f"Skill not found: {skill_id or name}",
+                }
+
+            # Use SkillUpdater to refresh from source
+            result = updater.update_skill(skill.id)
+
+            # Re-index skills if updated
+            if result.updated:
+                skills = storage.list_skills(
+                    project_id=project_id,
+                    limit=10000,
+                    include_global=True,
+                )
+                await search.index_skills_async(skills)
+
+            return {
+                "success": result.success,
+                "updated": result.updated,
+                "skipped": result.skipped,
+                "skip_reason": result.skip_reason,
+                "error": result.error,
+            }
+        except Exception as e:
+            return {
+                "success": False,
+                "error": str(e),
+            }
+
+    # --- install_skill tool ---
+
+    # Initialize loader
+    loader = SkillLoader()
+
+    @registry.tool(
+        name="install_skill",
+        description="Install a skill from a local path, GitHub URL, or ZIP archive. Auto-detects source type.",
+    )
+    async def install_skill(
+        source: str | None = None,
+        project_scoped: bool = False,
+    ) -> dict[str, Any]:
+        """
+        Install a skill from a source location.
+
+        Auto-detects source type:
+        - Local directory or SKILL.md file
+        - GitHub URL (owner/repo, github:owner/repo, https://github.com/...)
+        - ZIP archive (.zip file)
+
+        Args:
+            source: Path or URL to the skill source (required)
+            project_scoped: If True, install skill scoped to the project
+
+        Returns:
+            Dict with success status, skill_id, skill_name, and source_type
+        """
+        try:
+            # Validate input
+            if not source or not source.strip():
+                return {
+                    "success": False,
+                    "error": "source parameter is required",
+                }
+
+            source = source.strip()
+
+            # Determine source type and load skill
+            from gobby.storage.skills import SkillSourceType
+
+            parsed_skill = None
+            source_type: SkillSourceType | None = None
+
+            # Check if it's a GitHub URL/reference
+            # Pattern for owner/repo format (e.g., "anthropic/claude-code")
+            # Must match owner/repo pattern without path traversal or absolute paths
+            github_owner_repo_pattern = re.compile(
+                r"^[A-Za-z0-9_.-]+/[A-Za-z0-9_.-]+(/[A-Za-z0-9_./-]*)?$"
+            )
+
+            # Explicit GitHub references (always treated as GitHub, no filesystem check)
+            is_explicit_github = (
+                source.startswith("github:")
+                or source.startswith("https://github.com/")
+                or source.startswith("http://github.com/")
+            )
+
+            # For implicit owner/repo patterns, check local filesystem first
+            is_implicit_github_pattern = (
+                not is_explicit_github
+                and github_owner_repo_pattern.match(source)
+                and not source.startswith("/")
+                and ".." not in source  # Reject path traversal
+            )
+
+            # Determine if this is a GitHub reference:
+            # - Explicit refs are always GitHub
+            # - Implicit patterns are GitHub only if local path doesn't exist
+            is_github_ref = is_explicit_github or (
+                is_implicit_github_pattern and not Path(source).exists()
+            )
+            if is_github_ref:
+                # GitHub URL
+                try:
+                    parsed_skill = loader.load_from_github(source)
+                    source_type = "github"
+                except SkillLoadError as e:
+                    return {
+                        "success": False,
+                        "error": f"Failed to load from GitHub: {e}",
+                    }
+
+            # Check if it's a ZIP file
+            elif source.endswith(".zip"):
+                zip_path = Path(source)
+                if not zip_path.exists():
+                    return {
+                        "success": False,
+                        "error": f"ZIP file not found: {source}",
+                    }
+                try:
+                    parsed_skill = loader.load_from_zip(zip_path)
+                    source_type = "zip"
+                except SkillLoadError as e:
+                    return {
+                        "success": False,
+                        "error": f"Failed to load from ZIP: {e}",
+                    }
+
+            # Assume it's a local path
+            else:
+                local_path = Path(source)
+                if not local_path.exists():
+                    return {
+                        "success": False,
+                        "error": f"Path not found: {source}",
+                    }
+                try:
+                    parsed_skill = loader.load_skill(local_path)
+                    source_type = "local"
+                except SkillLoadError as e:
+                    return {
+                        "success": False,
+                        "error": f"Failed to load skill: {e}",
+                    }
+
+            if parsed_skill is None:
+                return {
+                    "success": False,
+                    "error": "Failed to load skill from source",
+                }
+
+            # Handle case where load_from_github/load_from_zip returns a list
+            if isinstance(parsed_skill, list):
+                if len(parsed_skill) == 0:
+                    return {
+                        "success": False,
+                        "error": "No skills found in source",
+                    }
+                # Use the first skill if multiple were found
+                parsed_skill = parsed_skill[0]
+
+            # Determine project ID for the skill
+            skill_project_id = project_id if project_scoped else None
+
+            # Store the skill
+            skill = storage.create_skill(
+                name=parsed_skill.name,
+                description=parsed_skill.description,
+                content=parsed_skill.content,
+                version=parsed_skill.version,
+                license=parsed_skill.license,
+                compatibility=parsed_skill.compatibility,
+                allowed_tools=parsed_skill.allowed_tools,
+                metadata=parsed_skill.metadata,
+                source_path=parsed_skill.source_path,
+                source_type=source_type,
+                source_ref=getattr(parsed_skill, "source_ref", None),
+                project_id=skill_project_id,
+                enabled=True,
+            )
+
+            # Re-index skills
+            skills = storage.list_skills(
+                project_id=project_id,
+                limit=10000,
+                include_global=True,
+            )
+            await search.index_skills_async(skills)
+
+            return {
+                "success": True,
+                "installed": True,
+                "skill_id": skill.id,
+                "skill_name": skill.name,
+                "source_type": source_type,
+            }
+        except Exception as e:
+            return {
+                "success": False,
+                "error": str(e),
+            }
+
+    return registry

gobby/mcp_proxy/tools/task_orchestration.py

@@ -11,6 +11,7 @@ from gobby.mcp_proxy.tools.orchestration.monitor import register_monitor
 from gobby.mcp_proxy.tools.orchestration.orchestrate import register_orchestrator
 from gobby.mcp_proxy.tools.orchestration.review import register_reviewer
 from gobby.mcp_proxy.tools.orchestration.utils import get_current_project_id
+from gobby.mcp_proxy.tools.orchestration.wait import register_wait
 
 if TYPE_CHECKING:
     from gobby.agents.runner import AgentRunner
@@ -74,4 +75,10 @@ def create_orchestration_registry(
         default_project_id=default_project_id,
     )
 
+    # Register wait tools
+    register_wait(
+        registry=registry,
+        task_manager=task_manager,
+    )
+
     return registry

gobby/mcp_proxy/tools/task_readiness.py

@@ -9,6 +9,7 @@ Provides tools for task readiness management:
 Extracted from tasks.py using Strangler Fig pattern for code decomposition.
 """
 
+import logging
 from collections.abc import Callable
 from typing import TYPE_CHECKING, Any
 
@@ -20,6 +21,8 @@ from gobby.workflows.state_manager import WorkflowStateManager
 if TYPE_CHECKING:
     from gobby.storage.tasks import LocalTaskManager
 
+logger = logging.getLogger(__name__)
+
 __all__ = [
     "create_readiness_registry",
     "is_descendant_of",
@@ -474,6 +477,16 @@ def create_readiness_registry(
         if best_proximity > 0:
             reasons.append("same branch as current work")
 
+        # Get recommended skills based on task category
+        recommended_skills: list[str] = []
+        try:
+            from gobby.workflows.context_actions import recommend_skills_for_task
+
+            task_brief = best_task.to_brief()
+            recommended_skills = recommend_skills_for_task(task_brief)
+        except Exception as e:
+            logger.debug(f"Skill recommendation failed: {e}")
+
         return {
             "suggestion": best_task.to_brief(),
             "score": best_score,
@@ -482,6 +495,7 @@ def create_readiness_registry(
                 {"ref": t.to_brief()["ref"], "title": t.title, "score": s}
                 for t, s, _, _ in scored[1:4]  # Show top 3 alternatives
             ],
+            "recommended_skills": recommended_skills,
         }
 
     registry.register(

gobby/mcp_proxy/tools/task_sync.py

@@ -159,7 +159,7 @@ def create_sync_registry(
 
     registry.register(
         name="link_commit",
-        description="Link a git commit to a task. Useful for tracking which commits implement a task.",
+        description="Link a git commit to a task. NOTE: For closing tasks, prefer close_task(task_id, commit_sha='...') which links and closes in one call. Use link_commit only when you need to link without closing.",
         input_schema={
             "type": "object",
             "properties": {