"""Base classes for AI-assistant integrations.

Provides:
- ``IntegrationOption`` — declares a CLI option an integration accepts.
- ``IntegrationBase`` — abstract base every integration must implement.
- ``MarkdownIntegration`` — concrete base for standard Markdown-format
  integrations (the common case — subclass, set three class attrs, done).
- ``TomlIntegration`` — concrete base for TOML-format integrations
  (Gemini, Tabnine — subclass, set three class attrs, done).
- ``SkillsIntegration`` — concrete base for integrations that install
  commands as agent skills (``speckit-<name>/SKILL.md`` layout).
"""

from __future__ import annotations

import re
import shutil
from abc import ABC
from dataclasses import dataclass
from pathlib import Path
from typing import TYPE_CHECKING, Any

if TYPE_CHECKING:
    from .manifest import IntegrationManifest


# ---------------------------------------------------------------------------
# IntegrationOption
# ---------------------------------------------------------------------------


@dataclass(frozen=True)
class IntegrationOption:
    """Declares an option that an integration accepts via ``--integration-options``.

    Attributes:
        name:      The flag name (e.g. ``"--commands-dir"``).
        is_flag:   ``True`` for boolean flags (``--skills``).
        required:  ``True`` if the option must be supplied.
        default:   Default value when not supplied (``None`` → no default).
        help:      One-line description shown in ``specify integrate info``.
    """

    name: str
    is_flag: bool = False
    required: bool = False
    default: Any = None
    help: str = ""


# ---------------------------------------------------------------------------
# IntegrationBase — abstract base class
# ---------------------------------------------------------------------------


class IntegrationBase(ABC):
    """Abstract base class every integration must implement.

    Subclasses must set the following class-level attributes:

    * ``key``              — unique identifier, matches actual CLI tool name
    * ``config``           — dict compatible with ``AGENT_CONFIG`` entries
    * ``registrar_config`` — dict compatible with ``CommandRegistrar.AGENT_CONFIGS``

    And may optionally set:

    * ``context_file``     — path (relative to project root) of the agent
                             context/instructions file (e.g. ``"CLAUDE.md"``)
    """

    # -- Must be set by every subclass ------------------------------------

    key: str = ""
    """Unique integration key — should match the actual CLI tool name."""

    config: dict[str, Any] | None = None
    """Metadata dict matching the ``AGENT_CONFIG`` shape."""

    registrar_config: dict[str, Any] | None = None
    """Registration dict matching ``CommandRegistrar.AGENT_CONFIGS`` shape."""

    # -- Optional ---------------------------------------------------------

    context_file: str | None = None
    """Relative path to the agent context file (e.g. ``CLAUDE.md``)."""

    invoke_separator: str = "."
    """Separator used in slash-command invocations (``"."`` → ``/speckit.plan``)."""

    # -- Markers for managed context section ------------------------------

    CONTEXT_MARKER_START = "<!-- SPECKIT START -->"
    CONTEXT_MARKER_END = "<!-- SPECKIT END -->"

    # -- Public API -------------------------------------------------------

    @classmethod
    def options(cls) -> list[IntegrationOption]:
        """Return options this integration accepts. Default: none."""
        return []

    def effective_invoke_separator(
        self, parsed_options: dict[str, Any] | None = None
    ) -> str:
        """Return the invoke separator for the given options.

        Subclasses whose separator depends on runtime options (e.g.
        Copilot in ``--skills`` mode) should override this method.
        The default implementation ignores *parsed_options* and returns
        the class-level ``invoke_separator``.
        """
        return self.invoke_separator

    def build_exec_args(
        self,
        prompt: str,
        *,
        model: str | None = None,
        output_json: bool = True,
    ) -> list[str] | None:
        """Build CLI arguments for non-interactive execution.

        Returns a list of command-line tokens that will execute *prompt*
        non-interactively using this integration's CLI tool, or ``None``
        if the integration does not support CLI dispatch.

        Subclasses for CLI-based integrations should override this.
        """
        return None

    def build_command_invocation(self, command_name: str, args: str = "") -> str:
        """Build the native slash-command invocation for a Spec Kit command.

        The CLI tools discover and execute commands from installed files
        on disk.  This method builds the invocation string the CLI
        expects — e.g. ``"/speckit.specify my-feature"`` for markdown
        agents or ``"/speckit-specify my-feature"`` for skills agents.

        *command_name* may be a full dotted name like
        ``"speckit.specify"``, an extension command like
        ``"speckit.git.commit"``, or a bare stem like ``"specify"``.
        """
        stem = command_name
        if stem.startswith("speckit."):
            stem = stem[len("speckit."):]

        invocation = f"/speckit.{stem}"
        if args:
            invocation = f"{invocation} {args}"
        return invocation

    def dispatch_command(
        self,
        command_name: str,
        args: str = "",
        *,
        project_root: Path | None = None,
        model: str | None = None,
        timeout: int = 600,
        stream: bool = True,
    ) -> dict[str, Any]:
        """Dispatch a Spec Kit command through this integration's CLI.

        By default this builds a slash-command invocation with
        ``build_command_invocation()`` and passes that prompt to
        ``build_exec_args()`` to construct the CLI command line.
        Integrations with custom dispatch behavior can override
        ``build_command_invocation()``, ``build_exec_args()``, or
        ``dispatch_command()`` directly.

        When *stream* is ``True`` (the default), stdout and stderr are
        piped directly to the terminal so the user sees live output.
        When ``False``, output is captured and returned in the dict.

        Returns a dict with ``exit_code``, ``stdout``, and ``stderr``.
        Raises ``NotImplementedError`` if the integration does not
        support CLI dispatch.
        """
        import subprocess

        prompt = self.build_command_invocation(command_name, args)
        # When streaming to the terminal, request text output so the
        # user sees readable output instead of raw JSONL events.
        exec_args = self.build_exec_args(
            prompt, model=model, output_json=not stream
        )

        if exec_args is None:
            msg = (
                f"Integration {self.key!r} does not support CLI dispatch. "
                f"Override build_exec_args() to enable it."
            )
            raise NotImplementedError(msg)

        cwd = str(project_root) if project_root else None

        if stream:
            # No timeout when streaming — the user sees live output and
            # can Ctrl+C at any time.  The timeout parameter is only
            # applied in the captured (non-streaming) branch below.
            try:
                result = subprocess.run(
                    exec_args,
                    text=True,
                    cwd=cwd,
                )
            except KeyboardInterrupt:
                return {
                    "exit_code": 130,
                    "stdout": "",
                    "stderr": "Interrupted by user",
                }
            return {
                "exit_code": result.returncode,
                "stdout": "",
                "stderr": "",
            }

        result = subprocess.run(
            exec_args,
            capture_output=True,
            text=True,
            cwd=cwd,
            timeout=timeout,
        )
        return {
            "exit_code": result.returncode,
            "stdout": result.stdout,
            "stderr": result.stderr,
        }

    # -- Primitives — building blocks for setup() -------------------------

    def shared_commands_dir(self) -> Path | None:
        """Return path to the shared command templates directory.

        Checks ``core_pack/commands/`` (wheel install) first, then
        ``templates/commands/`` (source checkout).  Returns ``None``
        if neither exists.
        """
        import inspect

        pkg_dir = Path(inspect.getfile(IntegrationBase)).resolve().parent.parent
        for candidate in [
            pkg_dir / "core_pack" / "commands",
            pkg_dir.parent.parent / "templates" / "commands",
        ]:
            if candidate.is_dir():
                return candidate
        return None

    def shared_templates_dir(self) -> Path | None:
        """Return path to the shared page templates directory.

        Contains ``vscode-settings.json``, ``spec-template.md``, etc.
        Checks ``core_pack/templates/`` then ``templates/``.
        """
        import inspect

        pkg_dir = Path(inspect.getfile(IntegrationBase)).resolve().parent.parent
        for candidate in [
            pkg_dir / "core_pack" / "templates",
            pkg_dir.parent.parent / "templates",
        ]:
            if candidate.is_dir():
                return candidate
        return None

    def list_command_templates(self) -> list[Path]:
        """Return sorted list of command template files from the shared directory."""
        cmd_dir = self.shared_commands_dir()
        if not cmd_dir or not cmd_dir.is_dir():
            return []
        return sorted(f for f in cmd_dir.iterdir() if f.is_file() and f.suffix == ".md")

    def command_filename(self, template_name: str) -> str:
        """Return the destination filename for a command template.

        *template_name* is the stem of the source file (e.g. ``"plan"``).
        Default: ``speckit.{template_name}.md``.  Subclasses override
        to change the extension or naming convention.
        """
        return f"speckit.{template_name}.md"

    def commands_dest(self, project_root: Path) -> Path:
        """Return the absolute path to the commands output directory.

        Derived from ``config["folder"]`` and ``config["commands_subdir"]``.
        Raises ``ValueError`` if ``config`` or ``folder`` is missing.
        """
        if not self.config:
            raise ValueError(
                f"{type(self).__name__}.config is not set; integration "
                "subclasses must define a non-empty 'config' mapping."
            )
        folder = self.config.get("folder")
        if not folder:
            raise ValueError(
                f"{type(self).__name__}.config is missing required 'folder' entry."
            )
        subdir = self.config.get("commands_subdir", "commands")
        return project_root / folder / subdir

    # -- File operations — granular primitives for setup() ----------------

    @staticmethod
    def copy_command_to_directory(
        src: Path,
        dest_dir: Path,
        filename: str,
    ) -> Path:
        """Copy a command template to *dest_dir* with the given *filename*.

        Creates *dest_dir* if needed.  Returns the absolute path of the
        written file.  The caller can post-process the file before
        recording it in the manifest.
        """
        dest_dir.mkdir(parents=True, exist_ok=True)
        dst = dest_dir / filename
        shutil.copy2(src, dst)
        return dst

    @staticmethod
    def record_file_in_manifest(
        file_path: Path,
        project_root: Path,
        manifest: IntegrationManifest,
    ) -> None:
        """Hash *file_path* and record it in *manifest*.

        *file_path* must be inside *project_root*.
        """
        rel = file_path.resolve().relative_to(project_root.resolve())
        manifest.record_existing(rel)

    @staticmethod
    def write_file_and_record(
        content: str,
        dest: Path,
        project_root: Path,
        manifest: IntegrationManifest,
    ) -> Path:
        """Write *content* to *dest*, hash it, and record in *manifest*.

        Creates parent directories as needed.  Writes bytes directly to
        avoid platform newline translation (CRLF on Windows).  Any
        ``\r\n`` sequences in *content* are normalised to ``\n`` before
        writing.  Returns *dest*.
        """
        dest.parent.mkdir(parents=True, exist_ok=True)
        normalized = content.replace("\r\n", "\n")
        dest.write_bytes(normalized.encode("utf-8"))
        rel = dest.resolve().relative_to(project_root.resolve())
        manifest.record_existing(rel)
        return dest

    def integration_scripts_dir(self) -> Path | None:
        """Return path to this integration's bundled ``scripts/`` directory.

        Looks for a ``scripts/`` sibling of the module that defines the
        concrete subclass (not ``IntegrationBase`` itself).
        Returns ``None`` if the directory doesn't exist.
        """
        import inspect

        cls_file = inspect.getfile(type(self))
        scripts = Path(cls_file).resolve().parent / "scripts"
        return scripts if scripts.is_dir() else None

    def install_scripts(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
    ) -> list[Path]:
        """Copy integration-specific scripts into the project.

        Copies files from this integration's ``scripts/`` directory to
        ``.specify/integrations/<key>/scripts/`` in the project.  Shell
        scripts are made executable.  All copied files are recorded in
        *manifest*.

        Returns the list of files created.
        """
        scripts_src = self.integration_scripts_dir()
        if not scripts_src:
            return []

        created: list[Path] = []
        scripts_dest = project_root / ".specify" / "integrations" / self.key / "scripts"
        scripts_dest.mkdir(parents=True, exist_ok=True)

        for src_script in sorted(scripts_src.iterdir()):
            if not src_script.is_file():
                continue
            dst_script = scripts_dest / src_script.name
            shutil.copy2(src_script, dst_script)
            if dst_script.suffix == ".sh":
                dst_script.chmod(dst_script.stat().st_mode | 0o111)
            self.record_file_in_manifest(dst_script, project_root, manifest)
            created.append(dst_script)

        return created

    # -- Agent context file management ------------------------------------

    @staticmethod
    def _ensure_mdc_frontmatter(content: str) -> str:
        """Ensure ``.mdc`` content has YAML frontmatter with ``alwaysApply: true``.

        If frontmatter is missing, prepend it.  If frontmatter exists but
        ``alwaysApply`` is absent or not ``true``, inject/fix it.

        Uses string/regex manipulation to preserve comments and formatting
        in existing frontmatter.
        """
        import re as _re

        leading_ws = len(content) - len(content.lstrip())
        leading = content[:leading_ws]
        stripped = content[leading_ws:]

        if not stripped.startswith("---"):
            return "---\nalwaysApply: true\n---\n\n" + content

        # Match frontmatter block: ---\n...\n---
        match = _re.match(
            r"^(---[ \t]*\r?\n)(.*?)(\r?\n---[ \t]*)(\r?\n|$)(.*)",
            stripped,
            _re.DOTALL,
        )
        if not match:
            return "---\nalwaysApply: true\n---\n\n" + content

        opening, fm_text, closing, sep, rest = match.groups()
        newline = "\r\n" if "\r\n" in opening else "\n"

        # Already correct?
        if _re.search(
            r"(?m)^[ \t]*alwaysApply[ \t]*:[ \t]*true[ \t]*(?:#.*)?$", fm_text
        ):
            return content

        # alwaysApply exists but wrong value — fix in place while preserving
        # indentation and any trailing inline comment.
        if _re.search(r"(?m)^[ \t]*alwaysApply[ \t]*:", fm_text):
            fm_text = _re.sub(
                r"(?m)^([ \t]*)alwaysApply[ \t]*:.*?([ \t]*(?:#.*)?)$",
                r"\1alwaysApply: true\2",
                fm_text,
                count=1,
            )
        elif fm_text.strip():
            fm_text = fm_text + newline + "alwaysApply: true"
        else:
            fm_text = "alwaysApply: true"

        return f"{leading}{opening}{fm_text}{closing}{sep}{rest}"

    @staticmethod
    def _build_context_section(plan_path: str = "") -> str:
        """Build the content for the managed section between markers.

        *plan_path* is the project-relative path to the current plan
        (e.g. ``"specs/<feature>/plan.md"``).  When empty, the section
        contains only the generic directive without a concrete path.
        """
        lines = [
            "For additional context about technologies to be used, project structure,",
            "shell commands, and other important information, read the current plan",
        ]
        if plan_path:
            lines.append(f"at {plan_path}")
        return "\n".join(lines)

    def upsert_context_section(
        self,
        project_root: Path,
        plan_path: str = "",
    ) -> Path | None:
        """Create or update the managed section in the agent context file.

        If the context file does not exist it is created with just the
        managed section.  If it exists, the content between
        ``<!-- SPECKIT START -->`` and ``<!-- SPECKIT END -->`` markers
        is replaced (or appended when no markers are found).

        Returns the path to the context file, or ``None`` when
        ``context_file`` is not set.
        """
        if not self.context_file:
            return None

        ctx_path = project_root / self.context_file
        section = (
            f"{self.CONTEXT_MARKER_START}\n"
            f"{self._build_context_section(plan_path)}\n"
            f"{self.CONTEXT_MARKER_END}\n"
        )

        if ctx_path.exists():
            content = ctx_path.read_text(encoding="utf-8-sig")
            start_idx = content.find(self.CONTEXT_MARKER_START)
            end_idx = content.find(
                self.CONTEXT_MARKER_END,
                start_idx if start_idx != -1 else 0,
            )

            if start_idx != -1 and end_idx != -1 and end_idx > start_idx:
                # Replace existing section (include the end marker + newline)
                end_of_marker = end_idx + len(self.CONTEXT_MARKER_END)
                # Consume trailing line ending (CRLF or LF)
                if end_of_marker < len(content) and content[end_of_marker] == "\r":
                    end_of_marker += 1
                if end_of_marker < len(content) and content[end_of_marker] == "\n":
                    end_of_marker += 1
                new_content = content[:start_idx] + section + content[end_of_marker:]
            elif start_idx != -1:
                # Corrupted: start marker without end — replace from start through EOF
                new_content = content[:start_idx] + section
            elif end_idx != -1:
                # Corrupted: end marker without start — replace BOF through end marker
                end_of_marker = end_idx + len(self.CONTEXT_MARKER_END)
                if end_of_marker < len(content) and content[end_of_marker] == "\r":
                    end_of_marker += 1
                if end_of_marker < len(content) and content[end_of_marker] == "\n":
                    end_of_marker += 1
                new_content = section + content[end_of_marker:]
            else:
                # No markers found — append
                if content:
                    if not content.endswith("\n"):
                        content += "\n"
                    new_content = content + "\n" + section
                else:
                    new_content = section

            # Ensure .mdc files have required YAML frontmatter
            if ctx_path.suffix == ".mdc":
                new_content = self._ensure_mdc_frontmatter(new_content)
        else:
            ctx_path.parent.mkdir(parents=True, exist_ok=True)
            # Cursor .mdc files require YAML frontmatter to be loaded
            if ctx_path.suffix == ".mdc":
                new_content = self._ensure_mdc_frontmatter(section)
            else:
                new_content = section

        normalized = new_content.replace("\r\n", "\n").replace("\r", "\n")
        ctx_path.write_bytes(normalized.encode("utf-8"))
        return ctx_path

    def remove_context_section(self, project_root: Path) -> bool:
        """Remove the managed section from the agent context file.

        Returns ``True`` if the section was found and removed.  If the
        file becomes empty (or whitespace-only) after removal it is
        deleted.
        """
        if not self.context_file:
            return False

        ctx_path = project_root / self.context_file
        if not ctx_path.exists():
            return False

        content = ctx_path.read_text(encoding="utf-8-sig")
        start_idx = content.find(self.CONTEXT_MARKER_START)
        end_idx = content.find(
            self.CONTEXT_MARKER_END,
            start_idx if start_idx != -1 else 0,
        )

        # Only remove a complete, well-ordered managed section. If either
        # marker is missing, leave the file unchanged to avoid deleting
        # unrelated user-authored content.
        if start_idx == -1 or end_idx == -1 or end_idx <= start_idx:
            return False

        removal_start = start_idx
        removal_end = end_idx + len(self.CONTEXT_MARKER_END)

        # Consume trailing line ending (CRLF or LF)
        if removal_end < len(content) and content[removal_end] == "\r":
            removal_end += 1
        if removal_end < len(content) and content[removal_end] == "\n":
            removal_end += 1

        # Also strip a blank line before the section if present
        if removal_start > 0 and content[removal_start - 1] == "\n":
            if removal_start > 1 and content[removal_start - 2] == "\n":
                removal_start -= 1

        new_content = content[:removal_start] + content[removal_end:]

        # Normalize line endings before comparisons
        normalized = new_content.replace("\r\n", "\n").replace("\r", "\n")

        # For .mdc files, treat Speckit-generated frontmatter-only content as empty
        if ctx_path.suffix == ".mdc":
            import re
            # Delete the file if only YAML frontmatter remains (no body content)
            frontmatter_only = re.match(
                r"^---\n.*?\n---\s*$", normalized, re.DOTALL
            )
            if not normalized.strip() or frontmatter_only:
                ctx_path.unlink()
                return True

        if not normalized.strip():
            ctx_path.unlink()
        else:
            ctx_path.write_bytes(normalized.encode("utf-8"))

        return True

    @staticmethod
    def resolve_command_refs(content: str, separator: str = ".") -> str:
        """Replace ``__SPECKIT_COMMAND_<NAME>__`` placeholders with invocations.

        Each placeholder encodes a command name in upper-case with
        underscores (e.g. ``__SPECKIT_COMMAND_PLAN__``,
        ``__SPECKIT_COMMAND_GIT_COMMIT__``).  The replacement uses
        *separator* to join the segments:

        * ``separator="."`` → ``/speckit.plan``, ``/speckit.git.commit``
        * ``separator="-"`` → ``/speckit-plan``, ``/speckit-git-commit``
        """
        return re.sub(
            r"__SPECKIT_COMMAND_([A-Z][A-Z0-9_]*)__",
            lambda m: "/speckit" + separator + m.group(1).lower().replace("_", separator),
            content,
        )

    @staticmethod
    def process_template(
        content: str,
        agent_name: str,
        script_type: str,
        arg_placeholder: str = "$ARGUMENTS",
        context_file: str = "",
        invoke_separator: str = ".",
    ) -> str:
        """Process a raw command template into agent-ready content.

        Performs the same transformations as the release script:
        1. Extract ``scripts.<script_type>`` value from YAML frontmatter
        2. Replace ``{SCRIPT}`` with the extracted script command
        3. Strip ``scripts:`` section from frontmatter
        4. Replace ``{ARGS}`` and ``$ARGUMENTS`` with *arg_placeholder*
        5. Replace ``__AGENT__`` with *agent_name*
        6. Replace ``__CONTEXT_FILE__`` with *context_file*
        7. Rewrite paths: ``scripts/`` → ``.specify/scripts/`` etc.
        8. Replace ``__SPECKIT_COMMAND_<NAME>__`` with invocation strings
        """
        # 1. Extract script command from frontmatter
        script_command = ""
        script_pattern = re.compile(
            rf"^\s*{re.escape(script_type)}:\s*(.+)$", re.MULTILINE
        )
        # Find the scripts: block
        in_scripts = False
        for line in content.splitlines():
            if line.strip() == "scripts:":
                in_scripts = True
                continue
            if in_scripts and line and not line[0].isspace():
                in_scripts = False
            if in_scripts:
                m = script_pattern.match(line)
                if m:
                    script_command = m.group(1).strip()
                    break

        # 2. Replace {SCRIPT}
        if script_command:
            content = content.replace("{SCRIPT}", script_command)

        # 3. Strip scripts: section from frontmatter
        lines = content.splitlines(keepends=True)
        output_lines: list[str] = []
        in_frontmatter = False
        skip_section = False
        dash_count = 0
        for line in lines:
            stripped = line.rstrip("\n\r")
            if stripped == "---":
                dash_count += 1
                if dash_count == 1:
                    in_frontmatter = True
                else:
                    in_frontmatter = False
                skip_section = False
                output_lines.append(line)
                continue
            if in_frontmatter:
                if stripped == "scripts:":
                    skip_section = True
                    continue
                if skip_section:
                    if line[0:1].isspace():
                        continue  # skip indented content under scripts
                    skip_section = False
            output_lines.append(line)
        content = "".join(output_lines)

        # 4. Replace {ARGS} and $ARGUMENTS
        content = content.replace("{ARGS}", arg_placeholder)
        content = content.replace("$ARGUMENTS", arg_placeholder)

        # 5. Replace __AGENT__
        content = content.replace("__AGENT__", agent_name)

        # 6. Replace __CONTEXT_FILE__
        content = content.replace("__CONTEXT_FILE__", context_file)

        # 7. Rewrite paths — delegate to the shared implementation in
        #    CommandRegistrar so extension-local paths are preserved and
        #    boundary rules stay consistent across the codebase.
        from specify_cli.agents import CommandRegistrar

        content = CommandRegistrar.rewrite_project_relative_paths(content)

        # 8. Replace __SPECKIT_COMMAND_<NAME>__ with invocation strings
        content = IntegrationBase.resolve_command_refs(content, invoke_separator)

        return content

    def setup(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
        parsed_options: dict[str, Any] | None = None,
        **opts: Any,
    ) -> list[Path]:
        """Install integration command files into *project_root*.

        Returns the list of files created.  Copies raw templates without
        processing.  Integrations that need placeholder replacement
        (e.g. ``{SCRIPT}``, ``__AGENT__``) should override ``setup()``
        and call ``process_template()`` in their own loop — see
        ``CopilotIntegration`` for an example.
        """
        templates = self.list_command_templates()
        if not templates:
            return []

        project_root_resolved = project_root.resolve()
        if manifest.project_root != project_root_resolved:
            raise ValueError(
                f"manifest.project_root ({manifest.project_root}) does not match "
                f"project_root ({project_root_resolved})"
            )

        dest = self.commands_dest(project_root).resolve()
        try:
            dest.relative_to(project_root_resolved)
        except ValueError as exc:
            raise ValueError(
                f"Integration destination {dest} escapes "
                f"project root {project_root_resolved}"
            ) from exc

        created: list[Path] = []

        for src_file in templates:
            dst_name = self.command_filename(src_file.stem)
            dst_file = self.copy_command_to_directory(src_file, dest, dst_name)
            self.record_file_in_manifest(dst_file, project_root, manifest)
            created.append(dst_file)

        # Upsert managed context section into the agent context file
        self.upsert_context_section(project_root)

        return created

    def teardown(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
        *,
        force: bool = False,
    ) -> tuple[list[Path], list[Path]]:
        """Uninstall integration files from *project_root*.

        Delegates to ``manifest.uninstall()`` which only removes files
        whose hash still matches the recorded value (unless *force*).
        Also removes the managed context section from the agent file.

        Returns ``(removed, skipped)`` file lists.
        """
        self.remove_context_section(project_root)
        return manifest.uninstall(project_root, force=force)

    # -- Convenience helpers for subclasses -------------------------------

    def install(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
        parsed_options: dict[str, Any] | None = None,
        **opts: Any,
    ) -> list[Path]:
        """High-level install — calls ``setup()`` and returns created files."""
        return self.setup(project_root, manifest, parsed_options=parsed_options, **opts)

    def uninstall(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
        *,
        force: bool = False,
    ) -> tuple[list[Path], list[Path]]:
        """High-level uninstall — calls ``teardown()``."""
        return self.teardown(project_root, manifest, force=force)


# ---------------------------------------------------------------------------
# MarkdownIntegration — covers ~20 standard agents
# ---------------------------------------------------------------------------


class MarkdownIntegration(IntegrationBase):
    """Concrete base for integrations that use standard Markdown commands.

    Subclasses only need to set ``key``, ``config``, ``registrar_config``
    (and optionally ``context_file``).  Everything else is inherited.

    ``setup()`` processes command templates (replacing ``{SCRIPT}``,
    ``{ARGS}``, ``__AGENT__``, rewriting paths) and upserts the
    managed context section into the agent context file.
    """

    def build_exec_args(
        self,
        prompt: str,
        *,
        model: str | None = None,
        output_json: bool = True,
    ) -> list[str] | None:
        if not self.config or not self.config.get("requires_cli"):
            return None
        args = [self.key, "-p", prompt]
        if model:
            args.extend(["--model", model])
        if output_json:
            args.extend(["--output-format", "json"])
        return args

    def setup(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
        parsed_options: dict[str, Any] | None = None,
        **opts: Any,
    ) -> list[Path]:
        templates = self.list_command_templates()
        if not templates:
            return []

        project_root_resolved = project_root.resolve()
        if manifest.project_root != project_root_resolved:
            raise ValueError(
                f"manifest.project_root ({manifest.project_root}) does not match "
                f"project_root ({project_root_resolved})"
            )

        dest = self.commands_dest(project_root).resolve()
        try:
            dest.relative_to(project_root_resolved)
        except ValueError as exc:
            raise ValueError(
                f"Integration destination {dest} escapes "
                f"project root {project_root_resolved}"
            ) from exc
        dest.mkdir(parents=True, exist_ok=True)

        script_type = opts.get("script_type", "sh")
        arg_placeholder = (
            self.registrar_config.get("args", "$ARGUMENTS")
            if self.registrar_config
            else "$ARGUMENTS"
        )
        created: list[Path] = []

        for src_file in templates:
            raw = src_file.read_text(encoding="utf-8")
            processed = self.process_template(
                raw, self.key, script_type, arg_placeholder,
                context_file=self.context_file or "",
            )
            dst_name = self.command_filename(src_file.stem)
            dst_file = self.write_file_and_record(
                processed, dest / dst_name, project_root, manifest
            )
            created.append(dst_file)

        # Upsert managed context section into the agent context file
        self.upsert_context_section(project_root)

        return created


# ---------------------------------------------------------------------------
# TomlIntegration — TOML-format agents (Gemini, Tabnine)
# ---------------------------------------------------------------------------


class TomlIntegration(IntegrationBase):
    """Concrete base for integrations that use TOML command format.

    Mirrors ``MarkdownIntegration`` closely: subclasses only need to set
    ``key``, ``config``, ``registrar_config`` (and optionally
    ``context_file``).  Everything else is inherited.

    ``setup()`` processes command templates through the same placeholder
    pipeline as ``MarkdownIntegration``, then converts the result to
    TOML format (``description`` key + ``prompt`` multiline string).
    """

    def build_exec_args(
        self,
        prompt: str,
        *,
        model: str | None = None,
        output_json: bool = True,
    ) -> list[str] | None:
        if not self.config or not self.config.get("requires_cli"):
            return None
        args = [self.key, "-p", prompt]
        if model:
            args.extend(["-m", model])
        if output_json:
            args.extend(["--output-format", "json"])
        return args

    def command_filename(self, template_name: str) -> str:
        """TOML commands use ``.toml`` extension."""
        return f"speckit.{template_name}.toml"

    @staticmethod
    def _extract_description(content: str) -> str:
        """Extract the ``description`` value from YAML frontmatter.

        Parses the YAML frontmatter so block scalar descriptions (``|``
        and ``>``) keep their YAML semantics instead of being treated as
        raw text.
        """
        import yaml

        frontmatter_text, _ = TomlIntegration._split_frontmatter(content)
        if not frontmatter_text:
            return ""
        try:
            frontmatter = yaml.safe_load(frontmatter_text) or {}
        except yaml.YAMLError:
            return ""

        if not isinstance(frontmatter, dict):
            return ""

        description = frontmatter.get("description", "")
        if isinstance(description, str):
            return description
        return ""

    @staticmethod
    def _split_frontmatter(content: str) -> tuple[str, str]:
        """Split YAML frontmatter from the remaining content.

        Returns ``("", content)`` when no complete frontmatter block is
        present. The body is preserved exactly as written so prompt text
        keeps its intended formatting.
        """
        if not content.startswith("---"):
            return "", content

        lines = content.splitlines(keepends=True)
        if not lines or lines[0].rstrip("\r\n") != "---":
            return "", content

        frontmatter_end = -1
        for i, line in enumerate(lines[1:], start=1):
            if line.rstrip("\r\n") == "---":
                frontmatter_end = i
                break

        if frontmatter_end == -1:
            return "", content

        frontmatter = "".join(lines[1:frontmatter_end])
        body = "".join(lines[frontmatter_end + 1 :])
        return frontmatter, body

    @staticmethod
    def _render_toml_string(value: str) -> str:
        """Render *value* as a TOML string literal.

        Uses a basic string for single-line values, multiline basic
        strings for values containing newlines, and falls back to a
        literal string or escaped basic string when delimiters appear in
        the content.
        """
        if "\n" not in value and "\r" not in value:
            escaped = value.replace("\\", "\\\\").replace('"', '\\"')
            return f'"{escaped}"'

        escaped = value.replace("\\", "\\\\")
        if '"""' not in escaped:
            if escaped.endswith('"'):
                return '"""\n' + escaped + '\\\n"""'
            return '"""\n' + escaped + '"""'
        if "'''" not in value and not value.endswith("'"):
            return "'''\n" + value + "'''"

        return (
            '"'
            + (
                value.replace("\\", "\\\\")
                .replace('"', '\\"')
                .replace("\n", "\\n")
                .replace("\r", "\\r")
                .replace("\t", "\\t")
            )
            + '"'
        )

    @staticmethod
    def _render_toml(description: str, body: str) -> str:
        """Render a TOML command file from description and body.

        Uses multiline basic strings (``\"\"\"``) with backslashes
        escaped, matching the output of the release script.  Falls back
        to multiline literal strings (``'''``) if the body contains
        ``\"\"\"``, then to an escaped basic string as a last resort.

        The body is ``rstrip("\\n")``'d before rendering, so the TOML
        value preserves content without forcing a trailing newline. As a
        result, multiline delimiters appear on their own line only when
        the rendered value itself ends with a newline.
        """
        toml_lines: list[str] = []

        if description:
            toml_lines.append(
                f"description = {TomlIntegration._render_toml_string(description)}"
            )
            toml_lines.append("")

        body = body.rstrip("\n")
        toml_lines.append(f"prompt = {TomlIntegration._render_toml_string(body)}")

        return "\n".join(toml_lines) + "\n"

    def setup(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
        parsed_options: dict[str, Any] | None = None,
        **opts: Any,
    ) -> list[Path]:
        templates = self.list_command_templates()
        if not templates:
            return []

        project_root_resolved = project_root.resolve()
        if manifest.project_root != project_root_resolved:
            raise ValueError(
                f"manifest.project_root ({manifest.project_root}) does not match "
                f"project_root ({project_root_resolved})"
            )

        dest = self.commands_dest(project_root).resolve()
        try:
            dest.relative_to(project_root_resolved)
        except ValueError as exc:
            raise ValueError(
                f"Integration destination {dest} escapes "
                f"project root {project_root_resolved}"
            ) from exc
        dest.mkdir(parents=True, exist_ok=True)

        script_type = opts.get("script_type", "sh")
        arg_placeholder = (
            self.registrar_config.get("args", "{{args}}")
            if self.registrar_config
            else "{{args}}"
        )
        created: list[Path] = []

        for src_file in templates:
            raw = src_file.read_text(encoding="utf-8")
            description = self._extract_description(raw)
            processed = self.process_template(
                raw, self.key, script_type, arg_placeholder,
                context_file=self.context_file or "",
            )
            _, body = self._split_frontmatter(processed)
            toml_content = self._render_toml(description, body)
            dst_name = self.command_filename(src_file.stem)
            dst_file = self.write_file_and_record(
                toml_content, dest / dst_name, project_root, manifest
            )
            created.append(dst_file)

        # Upsert managed context section into the agent context file
        self.upsert_context_section(project_root)

        return created


# ---------------------------------------------------------------------------
# YamlIntegration — YAML-format agents (Goose)
# ---------------------------------------------------------------------------


class YamlIntegration(IntegrationBase):
    """Concrete base for integrations that use YAML recipe format.

    Mirrors ``TomlIntegration`` closely: subclasses only need to set
    ``key``, ``config``, ``registrar_config`` (and optionally
    ``context_file``).  Everything else is inherited.

    ``setup()`` processes command templates through the same placeholder
    pipeline as ``MarkdownIntegration``, then converts the result to
    YAML recipe format (version, title, description, prompt block scalar).
    """

    def command_filename(self, template_name: str) -> str:
        """YAML commands use ``.yaml`` extension."""
        return f"speckit.{template_name}.yaml"

    @staticmethod
    def _extract_frontmatter(content: str) -> dict[str, Any]:
        """Extract frontmatter as a dict from YAML frontmatter block."""
        import yaml

        if not content.startswith("---"):
            return {}

        lines = content.splitlines(keepends=True)
        if not lines or lines[0].rstrip("\r\n") != "---":
            return {}

        frontmatter_end = -1
        for i, line in enumerate(lines[1:], start=1):
            if line.rstrip("\r\n") == "---":
                frontmatter_end = i
                break

        if frontmatter_end == -1:
            return {}

        frontmatter_text = "".join(lines[1:frontmatter_end])
        try:
            fm = yaml.safe_load(frontmatter_text) or {}
        except yaml.YAMLError:
            return {}

        return fm if isinstance(fm, dict) else {}

    @staticmethod
    def _split_frontmatter(content: str) -> tuple[str, str]:
        """Split YAML frontmatter from the remaining body content."""
        if not content.startswith("---"):
            return "", content

        lines = content.splitlines(keepends=True)
        if not lines or lines[0].rstrip("\r\n") != "---":
            return "", content

        frontmatter_end = -1
        for i, line in enumerate(lines[1:], start=1):
            if line.rstrip("\r\n") == "---":
                frontmatter_end = i
                break

        if frontmatter_end == -1:
            return "", content

        frontmatter = "".join(lines[1:frontmatter_end])
        body = "".join(lines[frontmatter_end + 1 :])
        return frontmatter, body

    @staticmethod
    def _human_title(identifier: str) -> str:
        """Convert an identifier to a human-readable title.

        Strips a leading ``speckit.`` prefix and replaces ``.``, ``-``,
        and ``_`` with spaces before title-casing.
        """
        text = identifier
        if text.startswith("speckit."):
            text = text[len("speckit.") :]
        return text.replace(".", " ").replace("-", " ").replace("_", " ").title()

    @staticmethod
    def _render_yaml(title: str, description: str, body: str, source_id: str) -> str:
        """Render a YAML recipe file from title, description, and body.

        Produces a Goose-compatible recipe with a literal block scalar
        for the prompt content.  Uses ``yaml.safe_dump()`` for the
        header fields to ensure proper escaping.
        """
        import yaml

        header = {
            "version": "1.0.0",
            "title": title,
            "description": description,
            "author": {"contact": "spec-kit"},
            "extensions": [{"type": "builtin", "name": "developer"}],
            "activities": ["Spec-Driven Development"],
        }

        header_yaml = yaml.safe_dump(
            header,
            sort_keys=False,
            allow_unicode=True,
            default_flow_style=False,
        ).strip()

        # Indent each line for YAML block scalar
        indented = "\n".join(f"  {line}" for line in body.split("\n"))

        lines = [header_yaml, "prompt: |", indented, "", f"# Source: {source_id}"]
        return "\n".join(lines) + "\n"

    def setup(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
        parsed_options: dict[str, Any] | None = None,
        **opts: Any,
    ) -> list[Path]:
        templates = self.list_command_templates()
        if not templates:
            return []

        project_root_resolved = project_root.resolve()
        if manifest.project_root != project_root_resolved:
            raise ValueError(
                f"manifest.project_root ({manifest.project_root}) does not match "
                f"project_root ({project_root_resolved})"
            )

        dest = self.commands_dest(project_root).resolve()
        try:
            dest.relative_to(project_root_resolved)
        except ValueError as exc:
            raise ValueError(
                f"Integration destination {dest} escapes "
                f"project root {project_root_resolved}"
            ) from exc
        dest.mkdir(parents=True, exist_ok=True)

        script_type = opts.get("script_type", "sh")
        arg_placeholder = (
            self.registrar_config.get("args", "{{args}}")
            if self.registrar_config
            else "{{args}}"
        )
        created: list[Path] = []

        for src_file in templates:
            raw = src_file.read_text(encoding="utf-8")
            fm = self._extract_frontmatter(raw)
            description = fm.get("description", "")
            if not isinstance(description, str):
                description = str(description) if description is not None else ""
            title = fm.get("title", "") or fm.get("name", "")
            if not isinstance(title, str):
                title = str(title) if title is not None else ""
            if not title:
                title = self._human_title(src_file.stem)

            processed = self.process_template(
                raw, self.key, script_type, arg_placeholder,
                context_file=self.context_file or "",
            )
            _, body = self._split_frontmatter(processed)
            yaml_content = self._render_yaml(
                title, description, body, f"templates/commands/{src_file.name}"
            )
            dst_name = self.command_filename(src_file.stem)
            dst_file = self.write_file_and_record(
                yaml_content, dest / dst_name, project_root, manifest
            )
            created.append(dst_file)

        # Upsert managed context section into the agent context file
        self.upsert_context_section(project_root)

        return created


# ---------------------------------------------------------------------------
# SkillsIntegration — skills-format agents (Codex, Kimi, Agy)
# ---------------------------------------------------------------------------


class SkillsIntegration(IntegrationBase):
    """Concrete base for integrations that install commands as agent skills.

    Skills use the ``speckit-<name>/SKILL.md`` directory layout following
    the `agentskills.io <https://agentskills.io/specification>`_ spec.

    Subclasses set ``key``, ``config``, ``registrar_config`` (and
    optionally ``context_file``) like any integration.  They may also
    override ``options()`` to declare additional CLI flags (e.g.
    ``--skills``, ``--migrate-legacy``).

    ``setup()`` processes each shared command template into a
    ``speckit-<name>/SKILL.md`` file with skills-oriented frontmatter.
    """

    invoke_separator = "-"

    def build_exec_args(
        self,
        prompt: str,
        *,
        model: str | None = None,
        output_json: bool = True,
    ) -> list[str] | None:
        if not self.config or not self.config.get("requires_cli"):
            return None
        args = [self.key, "-p", prompt]
        if model:
            args.extend(["--model", model])
        if output_json:
            args.extend(["--output-format", "json"])
        return args

    def skills_dest(self, project_root: Path) -> Path:
        """Return the absolute path to the skills output directory.

        Derived from ``config["folder"]`` and the configured
        ``commands_subdir`` (defaults to ``"skills"``).

        Raises ``ValueError`` when ``config`` or ``folder`` is missing.
        """
        if not self.config:
            raise ValueError(f"{type(self).__name__}.config is not set.")
        folder = self.config.get("folder")
        if not folder:
            raise ValueError(
                f"{type(self).__name__}.config is missing required 'folder' entry."
            )
        subdir = self.config.get("commands_subdir", "skills")
        return project_root / folder / subdir

    def build_command_invocation(self, command_name: str, args: str = "") -> str:
        """Skills use ``/speckit-<stem>`` (hyphenated directory name)."""
        stem = command_name
        if stem.startswith("speckit."):
            stem = stem[len("speckit."):]

        invocation = "/speckit-" + stem.replace(".", "-")
        if args:
            invocation = f"{invocation} {args}"
        return invocation

    def post_process_skill_content(self, content: str) -> str:
        """Post-process a SKILL.md file's content after generation.

        Called by external skill generators (presets, extensions) to let
        the integration inject agent-specific frontmatter or body
        transformations.  The default implementation returns *content*
        unchanged.  Subclasses may override — see ``ClaudeIntegration``.
        """
        return content

    def setup(
        self,
        project_root: Path,
        manifest: IntegrationManifest,
        parsed_options: dict[str, Any] | None = None,
        **opts: Any,
    ) -> list[Path]:
        """Install command templates as agent skills.

        Creates ``speckit-<name>/SKILL.md`` for each shared command
        template.  Each SKILL.md has normalised frontmatter containing
        ``name``, ``description``, ``compatibility``, and ``metadata``.
        """
        import yaml

        templates = self.list_command_templates()
        if not templates:
            return []

        project_root_resolved = project_root.resolve()
        if manifest.project_root != project_root_resolved:
            raise ValueError(
                f"manifest.project_root ({manifest.project_root}) does not match "
                f"project_root ({project_root_resolved})"
            )

        skills_dir = self.skills_dest(project_root).resolve()
        try:
            skills_dir.relative_to(project_root_resolved)
        except ValueError as exc:
            raise ValueError(
                f"Skills destination {skills_dir} escapes "
                f"project root {project_root_resolved}"
            ) from exc

        script_type = opts.get("script_type", "sh")
        arg_placeholder = (
            self.registrar_config.get("args", "$ARGUMENTS")
            if self.registrar_config
            else "$ARGUMENTS"
        )
        created: list[Path] = []

        for src_file in templates:
            raw = src_file.read_text(encoding="utf-8")

            # Derive the skill name from the template stem
            command_name = src_file.stem  # e.g. "plan"
            skill_name = f"speckit-{command_name.replace('.', '-')}"

            # Parse frontmatter for description
            frontmatter: dict[str, Any] = {}
            if raw.startswith("---"):
                parts = raw.split("---", 2)
                if len(parts) >= 3:
                    try:
                        fm = yaml.safe_load(parts[1])
                        if isinstance(fm, dict):
                            frontmatter = fm
                    except yaml.YAMLError:
                        pass

            # Process body through the standard template pipeline
            processed_body = self.process_template(
                raw, self.key, script_type, arg_placeholder,
                context_file=self.context_file or "",
                invoke_separator=self.invoke_separator,
            )
            # Strip the processed frontmatter — we rebuild it for skills.
            # Preserve leading whitespace in the body to match release ZIP
            # output byte-for-byte (the template body starts with \n after
            # the closing ---).
            if processed_body.startswith("---"):
                parts = processed_body.split("---", 2)
                if len(parts) >= 3:
                    processed_body = parts[2]

            # Select description — use the original template description
            # to stay byte-for-byte identical with release ZIP output.
            description = frontmatter.get("description", "")
            if not description:
                description = f"Spec Kit: {command_name} workflow"

            # Build SKILL.md with manually formatted frontmatter to match
            # the release packaging script output exactly (double-quoted
            # values, no yaml.safe_dump quoting differences).
            def _quote(v: str) -> str:
                escaped = v.replace("\\", "\\\\").replace('"', '\\"')
                return f'"{escaped}"'

            skill_content = (
                f"---\n"
                f"name: {_quote(skill_name)}\n"
                f"description: {_quote(description)}\n"
                f"compatibility: {_quote('Requires spec-kit project structure with .specify/ directory')}\n"
                f"metadata:\n"
                f"  author: {_quote('github-spec-kit')}\n"
                f"  source: {_quote('templates/commands/' + src_file.name)}\n"
                f"---\n"
                f"{processed_body}"
            )

            # Write speckit-<name>/SKILL.md
            skill_dir = skills_dir / skill_name
            skill_file = skill_dir / "SKILL.md"
            dst = self.write_file_and_record(
                skill_content, skill_file, project_root, manifest
            )
            created.append(dst)

        # Upsert managed context section into the agent context file
        self.upsert_context_section(project_root)

        return created