refactor: compile built-in prompts into binary instead of external files

This change moves built-in meta-prompts (save_prompt, update_prompt) from
external JSONL files into compiled Go code, making them always available
and version-controlled with the binary.

Changes:
- Add default_prompts.go with built-in meta-prompt definitions
- Update store to load built-ins from code, not files
- Add protection: built-ins cannot be updated/deleted
- Handle name conflicts: built-ins take precedence with warnings
- Update docs to reflect new architecture (no default.jsonl needed)
- Add comprehensive tests for built-in protection
- Add hexai-mcp-server binary to .gitignore

Benefits:
- Built-ins always in sync with binary version
- No setup required (no default.jsonl to manage)
- Clear separation between built-in and user prompts
- Protection prevents accidental modification of meta-prompts

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

1 files changed, 149 insertions, 0 deletions

diff --git a/internal/promptstore/default_prompts.go b/internal/promptstore/default_prompts.go
new file mode 100644
index 0000000..ea8c793
--- /dev/null
+++ b/internal/promptstore/default_prompts.go
@@ -0,0 +1,149 @@
+// Summary: Built-in meta-prompts for prompt management.
+package promptstore
+
+import (
+	"time"
+)
+
+// DefaultPrompts returns the built-in meta-prompts for prompt management.
+// These prompts help users create and update prompts interactively using Claude's
+// access to conversation context.
+func DefaultPrompts() []Prompt {
+	now := time.Now()
+
+	return []Prompt{
+		{
+			Name:        "save_prompt",
+			Title:       "Save Current Conversation as Prompt",
+			Description: "Interactively create a new prompt template from the current conversation. Claude will analyze the conversation, ask clarifying questions about templating, show a preview, and wait for approval before saving.",
+			Arguments: []PromptArgument{
+				{
+					Name:        "prompt_name",
+					Description: "Unique identifier for the new prompt (lowercase, underscores allowed)",
+					Required:    true,
+				},
+				{
+					Name:        "prompt_title",
+					Description: "Human-readable display name for the new prompt",
+					Required:    true,
+				},
+			},
+			Messages: []PromptMessage{
+				{
+					Role: "user",
+					Content: MessageContent{
+						Type: "text",
+						Text: `I want to create a new prompt template named '{{prompt_name}}' with title '{{prompt_title}}'.
+
+Please help me by:
+1) Analyzing our current conversation to understand what should be templated
+2) Asking me clarifying questions about:
+   - What parts should be template arguments vs fixed text
+   - What description would best explain this prompt's purpose
+   - What tags would help categorize it
+   - Whether multi-turn messages are needed
+3) Showing me a complete preview of the prompt structure in a code block
+4) Only after I approve, use the MCP prompts/create method to save it
+
+IMPORTANT FORMATTING RULES for clarifying questions:
+- Use numbered questions: 1), 2), 3)
+- ANY CHOICE MUST BE NUMBERED using combined format: 1a), 1b), 1c), 2a), 2b), etc.
+- NEVER use standalone letters like "a)" - always combine with question number
+- NEVER use dashes (-) or bullets (•) for options
+- Every option must be numbered for easy selection by the user
+
+Examples:
+  1) Question Category
+  Which do you prefer?
+  1a) First option
+  1b) Second option
+  1c) Third option
+
+  2) Arguments
+  Should this accept parameters?
+  2a) No arguments - fixed behavior
+  2b) Optional file_pattern argument
+  2c) Multiple optional arguments
+
+  3) Sub-items
+  Consider these aspects:
+  3a) First aspect to consider
+  3b) Second aspect to consider
+  3c) Third aspect to consider
+
+  4) Multiple sub-questions
+  4a) Sub-question one?
+      Answer options here
+  4b) Sub-question two?
+      Answer options here
+
+Start by examining our conversation and asking your clarifying questions using this format.`,
+					},
+				},
+			},
+			Tags:    []string{"meta", "prompt-management", "interactive"},
+			Created: now,
+			Updated: now,
+		},
+		{
+			Name:        "update_prompt",
+			Title:       "Update Existing Prompt",
+			Description: "Interactively modify an existing prompt. Claude will fetch the current version, ask what changes you want, show a preview with changes highlighted, and wait for approval before updating.",
+			Arguments: []PromptArgument{
+				{
+					Name:        "prompt_name",
+					Description: "Name of the existing prompt to update",
+					Required:    true,
+				},
+			},
+			Messages: []PromptMessage{
+				{
+					Role: "user",
+					Content: MessageContent{
+						Type: "text",
+						Text: `I want to update the existing prompt '{{prompt_name}}'.
+
+Please help me by:
+1) First, retrieve the current prompt using prompts/get to show me what exists
+2) Ask me what changes I want to make (description, arguments, messages, tags)
+3) If I reference content from our current conversation, help extract and template it
+4) Show me a complete preview of the updated prompt with changes highlighted
+5) Only after I approve, use the MCP prompts/update method to save the changes
+
+IMPORTANT FORMATTING RULES for clarifying questions:
+- Use numbered questions: 1), 2), 3)
+- ANY CHOICE MUST BE NUMBERED using combined format: 1a), 1b), 1c), 2a), 2b), etc.
+- NEVER use standalone letters like "a)" - always combine with question number
+- NEVER use dashes (-) or bullets (•) for options
+- Every option must be numbered for easy selection by the user
+
+Examples:
+  1) Question Category
+  Which do you prefer?
+  1a) First option
+  1b) Second option
+  1c) Third option
+
+  2) Changes to Make
+  What aspects should be updated?
+  2a) Description only
+  2b) Arguments only
+  2c) Both description and arguments
+  2d) Complete rewrite
+
+  3) Multiple aspects
+  Consider:
+  3a) First aspect to evaluate
+  3b) Second aspect to evaluate
+  3c) Third aspect to evaluate
+
+Start by fetching and showing me the current prompt, then ask clarifying questions using this format.`,
+					},
+				},
+			},
+			Tags:    []string{"meta", "prompt-management", "interactive"},
+			Created: now,
+			Updated: now,
+		},
+	}
+}