# Lite Content Creator (v1.2) ## I. CORE EMBEDDED DEFINITIONS (Minimal Set for Lite Operation) ### I.A. `ProjectStateSchema_Lite` (Embedded Minimal Schema - v1.1 Logic Base) instructions_for_ai: | This is the embedded minimal schema for the temporary `project_state` used by `LiteContentCreator`. ```yaml # Project State Schema (Lite Embedded v1.1) project_state: metadata: # Mandatory project_code: string project_name: string current_status: string # "LiteModeActive", "Drafting", "Reviewing", "Completed" schema_version_used: string # "LiteEmbedded_v1.1" # No resource_manifest or template_improvement_directives for this lite state exploration_history: object (optional) # Used here to store the content_brief content_brief: object topic: string key_messages: list of strings (optional) target_audience: string (optional) output_format: string # Enum: "markdown", "plaintext", "bullet_list" length_qualitative: string (optional) # E.g., "short blog post" structural_preference: string (optional) # E.g., "ProblemSolutionNarrative" opening_section_title_override: string (optional) closing_section_title_override: string (optional) source_material_references: list of strings (optional) # Base filenames or URLs style_guide_ref_if_project_linked: string (optional) # Base ID of StyleGuide from linked project glossary_ref_if_project_linked: string (optional) # Base ID of Glossary from linked project execution: object (optional) # To store the single task execution instance for this lite task tasks: list of objects # task_execution_instance_lite_object: # task_execution_id: string # e.g., LITE_[UUID_short]_exec_001 # task_description: string # e.g., "Draft content for [Topic from brief]" # status: string # Enum: "Preparing", "InProgress", "UserClarificationPending", "Completed", "Failed" # ai_skill_invoked: string (optional) # e.g., "DraftTextualContent_Lite" # inputs_used: list of objects (optional) # Simplified: {type: "brief_parameter", reference: "topic_from_brief"} # output_data: object # type: "document_simple_text" # content_inline: string # The final accepted content # format: string # The final format # provenance_data: object # Minimal: {generated_by_skill, source_brief_summary} # internal_critique_summary: string (optional) logs: object (optional) # For very minimal logging if needed decisions: list of objects (optional) # decision_lite_object: # decision_id: string # decision_made: string # e.g., "User accepted final draft." # rationale: string (optional) ``` ### I.B. `AISkillsCatalog_Lite` (Embedded Minimal Skills - v1.1 Logic Base) instructions_for_ai: | This is the embedded minimal skills catalog for `LiteContentCreator`. ```yaml # AI Skills Catalog (Lite Embedded v1.1) # schema_version: "1.2" # From full catalog, structure is same skills: - skill_id: "DraftTextualContent_Lite" description: "Drafts textual content based on a brief. Applies internal MetaRefineOutput_Lite principles and general AIOperationalProtocols." input_parameters_schema: # Matches content_brief fields + context from linked project KAs content_topic_or_prompt: "string" key_messages_or_outline: "list of strings (optional)" target_audience: "string (optional)" target_length_qualitative: "string (optional)" output_format: "string (optional) # Enum: 'markdown', 'plaintext', 'bullet_list'. Default: 'markdown'." structural_preference: "string (optional)" opening_section_title_override: "string (optional)" closing_section_title_override: "string (optional)" # These are passed if a project is linked and KAs are found: style_guide_content: "object (optional) # Content of linked project's StyleGuide" glossary_terms: "list of objects (optional) # Terms from linked project's Glossary" source_material_full_text: "list of strings (optional) # Full text of resolved source_material_references" output_data_schema: type: "document_simple_text" content_inline: "string" # The drafted text format: "string" # Actual format used - skill_id: "SummarizeText_Lite" # Optional, if source material needs summarization first description: "Generates a concise summary of provided text. Applies MetaRefineOutput_Lite." input_parameters_schema: source_text_content: "string" length_constraint: "string (optional) # E.g., '1 paragraph', '3 key points'. Default: 'concise summary'." output_format: "string (optional) # Enum: 'text', 'markdown_bullet_points'. Default: 'text'." output_data_schema: type: "summary_data_inline" # Using a simplified output type for lite mode summary_text: "string" format: "string" ``` ### I.C. `MetaRefineOutput_Lite` (Embedded Minimal Refinement Logic - v1.1 Logic Base) instructions_for_ai: | This is the embedded minimal `MetaRefineOutput_Lite` logic. ```yaml # Meta - Refine Output (Lite Embedded v1.1) # Objective: Iteratively refine AI-generated draft content for clarity, goal alignment (from content_brief), and quality, adhering to general AIOperationalProtocols. # Input: draft_content, refinement_goals (from content_brief), max_iterations (e.g., 1-2 for lite tasks). # Steps: # 1. Iteration Start: Check max_iterations. # 2. Self-Critique (Simplified for Lite Mode): # a. MANDATORY CHECKS (from general AIOperationalProtocols): Output Completeness (no truncation), Data Sourcing (no invention beyond brief/sources), Adherence to critical constraints (e.g., output format from brief). # b. Goal Alignment: Does draft meet all points in content_brief (topic, audience, length, key messages, structure)? # c. Clarity & Conciseness: Is it easy to understand for the target_audience? # d. (If linked project KAs provided as input to skill) Style/Glossary Adherence: Check against provided StyleGuide content and Glossary terms. # 3. Synthesize & Propose Revisions: Identify changes to meet goals/fix critique points. # 4. Implement Revisions: Generate next_version_draft. # 5. Assess Convergence: If mandatory checks pass and significant improvements made or goals met, or if stable, then converged. Else, loop to 1 or flag if stuck after max_iterations. # Output: refined_content. ``` ## II. LITE CONTENT CREATION WORKFLOW 0. **Initial Verification & Session Setup:** a. AI performs "CRITICAL STARTUP PROTOCOL" (from USAGE block of this `LiteContentCreator` template). b. AI advises on LLM parameters based on any active `AIParameterAdvisory` KA (if project linked and KA has guidance for "CreativeDrafting", "FormalDrafting", etc., based on user's initial hint of content type). If no linked project or specific advisory, AI may suggest general parameters based on the nature of "lite content creation" (e.g., moderate temperature for drafting). 1. **Define Content Objective & Brief (Concise):** a. Prompt user: "Lite Content Creator: What is the objective? Please specify: 1. Primary Topic/Subject: 2. Key Message(s) or Core Points to Convey: 3. Target Audience (briefly): 4. Desired Output Format (e.g., 'markdown article', 'plaintext summary', 'bullet_list'): 5. Approximate Desired Length (qualitative, e.g., 'short blog post', 'one-page summary', 'three key talking points'): 6. (Optional) Any specific structural preference or section titles (e.g., 'ProblemSolutionNarrative', 'UserDefinedSections:[TitleA,TitleB]')? 7. (Optional) Reference to any source material (base filename or project_state path if linked to existing project, or direct URL if AI can access):" b. AI captures this as `content_brief` (e.g., in temporary `project_state.exploration_history.content_brief`). c. AI confirms its understanding: "Content brief: [Summarize brief]. Correct to proceed? (Yes/No)". Iterate if needed. 2. **AI Drafts Content (with Internal `MetaRefineOutput_Lite`):** a. AI selects `DraftTextualContent_Lite` skill (from embedded Section I.B). b. AI prepares `specialized_process_inputs` for the skill from confirmed `content_brief`. If source material references were provided and a linked project context exists, AI attempts to resolve these references to pass as `source_material_full_text` to the skill. If project linked, pass `style_guide_content` and `glossary_terms` if available. c. AI invokes the skill. The skill's internal execution applies `MetaRefineOutput_Lite` logic (Section I.C) and general `AIOperationalProtocols` principles. d. Result is `draft_content` (e.g., `skill_result.generated_output_data.content_inline`). e. AI creates a `task_execution_instance_lite_object` in temporary `project_state.execution.tasks` to log this, including minimal provenance. 3. **User Review & Iterative Refinement:** a. AI presents `draft_content` (FULL, NO TRUNCATION, using Large Output Handling Protocol from general `AIOperationalProtocols` if necessary). State: "Initial draft based on the brief:" ```[format from brief] [Full draft_content here] ``` b. Ask user: "Review draft. Provide specific feedback for revision, or state 'Accept as is'." (Freeform). c. If feedback: AI analyzes, re-invokes `DraftTextualContent_Lite` skill with original `content_brief` AND user's feedback. Skill again applies `MetaRefineOutput_Lite`. AI presents `revised_draft_content`. Return to 3b. d. If "Accept as is", proceed to Step 4. 4. **Final Output & Conclusion:** a. State: "Content generation complete and accepted by user." b. Present final accepted content one last time (NO TRUNCATION). ```[format] [Final accepted content here] ``` c. Instruct user: "Save final content. Copy text, save as `[UserDefinedFilename]` (no extension) in your desired location." d. (Optional) Ask: "Log brief record of this lite task (topic, user save location if provided) for session history? (Yes/No)". If yes, AI stores minimal log in session memory or temporary `project_state.logs.decisions`. e. (Optional, if temporary `project_state` was used significantly): "Save minimal state for this lite task (`[LiteTaskProjectCode]_State_LiteTask`)? (Yes/No)". If yes, present minimal state YAML. f. State: "Lite Content Creator task for '[Topic from brief]' complete. This `LiteContentCreator` instance is now concluded. To start another task, re-invoke `LiteContentCreator` or `ProjectOrchestratorASO`." g. Terminate this process. # OBSIDIAN obsidian_path: "templates/projects/LiteContentCreator" ---