[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_153":154,"_155":156},"routes/_app.x.$handle.$name",{"_95":94,"_32":45,"_103":85,"_27":104,"_13":14,"_105":-5,"_11":106,"_107":108,"_109":110,"_111":112,"_113":114,"_20":-5,"_115":-5,"_116":-5,"_117":25,"_118":119,"_120":57,"_121":-5,"_122":-5,"_123":124},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":-5,"_18":-5},"project",{"_43":44,"_32":45,"_38":46,"_47":-5,"_48":25,"_49":50,"_51":50,"_52":-5,"_53":-5,"_18":54,"_55":31,"_56":57,"_58":57,"_59":-5,"_60":61,"_62":63,"_64":-5,"_65":-5,"_66":-5,"_67":-5,"_68":69,"_70":69,"_71":72,"_73":72,"_74":72,"_75":69,"_76":69,"_77":78,"_79":80,"_81":82,"_83":25},"version",5,"changes",[19],"aiChatInfo","userId",{"_20":21,"_22":23,"_24":25,"_26":25,"_27":28,"_29":28},"file",{"_30":31,"_32":28,"_33":34,"_35":37,"_38":39,"_40":14,"_41":42,"_24":25},"previous",{"_30":31,"_32":28,"_33":34,"_35":36},"deleted",false,"deletedHere","path","main.ts","latestPath","fileId","fe8bba12-a5c0-11f0-8fdf-0224a6c84d84","name","fileType","http","content","import { Hono } from \"npm:hono@4.4.12\";\n// @ts-ignore\nimport { OpenAI } from \"https://esm.town/v/std/openai@v4\";\n\n// --- 1. TYPES (Unchanged) ---\ninterface ConfluenceOutput {\n section_title_a: string;\n revised_prompt_text: string;\n section_title_b: string;\n questions: string[];\n}\ninterface DomainTransfer {\n domain: string;\n rigorConstraint: string;\n reasoning: string;\n}\n\n// --- 2. PROMPTS (Unchanged) ---\nconst getBaseSystemPrompt = (\n userGoal: string,\n transferFunction: DomainTransfer,\n): string => `\nYou are the ultimate System Prompt Guy, an Elite Prompt Engineering Confluence model, tasked with collaboratively refining a user's initial concept into the most resilient, precise, and effective final system prompt through an iterative, Test-Time Computation (TTC)-inspired design process.\n\nThe user's core task/goal for their prompt is: \"${userGoal}\"\n\n---\n# INJECTED LOGICAL RIGOR: Domain Transfer Function (DTF)\nThe task has been mapped to the high-rigor domain of **${transferFunction.domain}**.\nYou MUST ensure that the following novel constraint is integrated into the \"Prompt Engineering Constraints\" section of the Revised Prompt:\n- **DTF Rigor Injection:** You must integrate the following constraint: *${transferFunction.rigorConstraint}*.\n---\n\n# Instructions and Core Role\nYour core role is to serve as the system prompt engineer, applying a rigorous set of structural and logical constraints to iteratively draft and refine a Revised Prompt based on the user's concept. You must prioritize Reasoning Before Conclusions in every step of your process.\n\n---\n# Prompt Engineering Constraints\nWhen drafting the [REVISED_PROMPT_TEXT], you must ensure it rigorously adheres to the following engineering standards:\n\n1. **Direct Start (Yes):** Start with a concise, single-sentence instruction line defining the model's role and primary task.\n2. **Formatting (Yes):** Use markdown headings (#, ##) and structure for maximum clarity.\n3. **Separators (Yes):** Use a horizontal line (---) immediately preceding each main section heading.\n4. **Reasoning Order (Yes):** Explicitly enforce that the executing model must provide its reasoning/justification BEFORE the final conclusion or classification.\n5. **Clarity & Voice (Yes):** Use clear, specific, active voice language; avoid ambiguity and unnecessary statements.\n6. **Completeness (Yes):** Include all necessary sections: Instructions, Steps, and Output Format.\n7. **Preservation (Yes):** Retain all specific details, guidelines, variables, and constants from the user's input.\n8. **Structured Bias (Yes):** Bias toward JSON for any final output that involves classification, entity extraction, or structured data.\n9. **DTF Rigor Injection (Yes):** *${transferFunction.rigorConstraint}*\n\n---\n# Steps: Iterative Refinement Process\nExecute the following steps sequentially for every user turn to guarantee Adaptive Compute Allocation and 100% instruction adherence.\n\n1. **Core Analysis & Drafting (TTC Reasoning):** Based on the user's latest input, draft the [REVISED_PROMPT_TEXT] by rigorously applying all constraints from the **Prompt Engineering Constraints** section.\n2. **Identify Gaps (Reasoning):** Analyze the initial input and the drafted [REVISED_PROMPT_TEXT] to determine what specific, additional information is required to maximize the prompt's precision.\n3. **Formulate Questions (Conclusion):** Based on the gaps identified, formulate a list of specific, highly relevant questions to ask the user.\n4. **Generate Output:** Present the output in the exact JSON format specified in the **Output Format Specification** section.\n\n---\n# Output Format Specification\nYour final, iterative output must be a single, valid JSON object. DO NOT include any text, commentary, or markdown outside of the JSON block.\n\n\\`\\`\\`json\n{\n \"section_title_a\": \"Revised Prompt\",\n \"revised_prompt_text\": \"[REVISED_PROMPT_TEXT]\",\n \"section_title_b\": \"Questions\",\n \"questions\": [\n \"[QUESTION_1]\",\n \"[QUESTION_2]\"\n ]\n}\n\\`\\`\\`\n`;\nconst getDtfPrompt = (userGoal: string): string => `\nYou are the Domain Transfer Agent (DTA). Your sole purpose is to increase the rigor of a given professional task by mapping its core logical demands to a high-stakes, highly disciplined professional domain.\n---\n# Instructions\n1. **Analyze the User Goal:** Identify the core logical functions required (e.g., speed, accuracy, risk assessment, sequential protocol, data synthesis).\n2. **Identify a High-Rigor Domain:** Select a domain (e.g., aerospace, surgical medicine, military protocol, financial regulation) that shares this logical function but operates under extreme stress or constraint.\n3. **Synthesize a Rigor Constraint:** Create a single, concise, and highly demanding constraint based on the selected domain's *modus operandi*.\n4. **Output Format:** Respond STRICTLY with a single, minified JSON object.\n---\n# Output Format Specification\n\\`\\`\\`json\n{\n \"domain\": \"[Selected High-Rigor Domain]\",\n \"rigorConstraint\": \"[The New Prompt Constraint to Inject]\",\n \"reasoning\": \"[Brief justification for the domain mapping]\"\n}\n\\`\\`\\`\n---\n# User Goal\nThe professional task the user wants to accomplish with their prompt is: \"${userGoal}\"\n`;\n\n// --- 3. AI LOGIC (Unchanged) ---\nasync function getDomainTransfer(\n openai: OpenAI,\n userGoal: string,\n model: string,\n): Promise {\n try {\n const completion = await openai.chat.completions.create({\n model: model,\n messages: [{ role: \"system\", content: getDtfPrompt(userGoal) }],\n response_format: { type: \"json_object\" },\n });\n const result: DomainTransfer = JSON.parse(\n completion.choices[0].message.content,\n );\n if (!result.domain || !result.rigorConstraint || !result.reasoning) {\n throw new Error(\n \"DTF response missing essential keys.\",\n );\n }\n console.log(\n `DTF Log - Domain: ${result.domain}, Reasoning: ${result.reasoning}`,\n );\n return result;\n } catch (e) {\n console.error(\"DTF Execution or Parsing Error:\", e);\n return {\n domain: \"Nuclear Reactor Protocol\",\n rigorConstraint: \"All output must be auditable and traceable.\",\n reasoning: \"Fallback due to AI parsing error.\",\n };\n }\n}\nasync function runPromptRefinement(\n openai: OpenAI,\n userGoal: string,\n model: string,\n): Promise {\n const transferFunction = await getDomainTransfer(openai, userGoal, model);\n const eliteSystemPrompt = getBaseSystemPrompt(userGoal, transferFunction);\n const completion = await openai.chat.completions.create({\n model: model,\n messages: [\n { role: \"system\", content: eliteSystemPrompt },\n {\n role: \"user\",\n content: `My initial prompt concept is: \"${userGoal}\". Begin.`,\n },\n ],\n response_format: { type: \"json_object\" },\n });\n const refinedOutput: ConfluenceOutput = JSON.parse(\n completion.choices[0].message.content,\n );\n return { ...refinedOutput, domainTransfer: transferFunction };\n}\n\n// --- 4. HTML GENERATION (Overhauled) ---\nfunction generateHTML(sourceUrl: string): string {\n // All CSS, HTML, and client-side JS are now inside this single template literal.\n // The sourceUrl is safely injected directly into the script.\n return `\n \n \n \n \n \n Faerie Prompt Confluence 🌿✨\n \n \n \n \n \n \n

\n \n

Faerie Prompt Confluence 🌿✨

Weaving logical rigor from high-stakes domains into resilient, precise, and enchanted system prompts for your AI companions.

\n \n

Whisper Your Intent

\n \n

Channeling Domain Spirit...

\n Enchanted Prompt\n \n

Domain Spirit 🌲

Domain:

Reasoning:

Injected Constraint:

\n \n

Further Questions

\n \n \n \n `;\n}\n\n// --- 5. SERVER (Unchanged) ---\nconst app = new Hono();\n\n// The GET handler calculates the URL, generates the HTML string, and sends it.\napp.get(\"/\", (c) => {\n const url = new URL(c.req.url);\n const sourceUrl = `https://${url.hostname}${url.pathname}`;\n const fullSourceUrl = sourceUrl.endsWith(\"/\") ? sourceUrl : `${sourceUrl}/`;\n\n return c.html(generateHTML(fullSourceUrl));\n});\n\n// The POST handler for the API remains unchanged.\napp.post(\"/refine\", async (c) => {\n try {\n const { userGoal, model } = await c.req.json();\n const allowedModels = [\"gpt-4o\", \"gpt-4-turbo\", \"gpt-3.5-turbo\"];\n\n if (\n !userGoal || typeof userGoal !== \"string\" || userGoal.trim().length < 10\n ) {\n return c.json({ error: \"The user goal must be a valid string.\" }, 400);\n }\n if (!model || !allowedModels.includes(model)) {\n return c.json({ error: \"A valid AI model must be selected.\" }, 400);\n }\n\n const openai = new OpenAI();\n const result = await runPromptRefinement(openai, userGoal, model);\n return c.json(result);\n } catch (error) {\n console.error(\"Critical Error in /refine:\", error);\n return c.json(\n { error: \"Server processing error.\", reason: error.message },\n 500,\n );\n }\n});\n\nexport default app.fetch;","import { Hono } from \"npm:hono@4.4.12\";\n// @ts-ignore\nimport { OpenAI } from \"https://esm.town/v/std/openai\";\n\n// --- 1. TYPES (Unchanged) ---\ninterface ConfluenceOutput {\n section_title_a: string;\n revised_prompt_text: string;\n section_title_b: string;\n questions: string[];\n}\ninterface DomainTransfer {\n domain: string;\n rigorConstraint: string;\n reasoning: string;\n}\n\n// --- 2. PROMPTS (Unchanged) ---\nconst getBaseSystemPrompt = (\n userGoal: string,\n transferFunction: DomainTransfer,\n): string => `\nYou are the ultimate System Prompt Guy, an Elite Prompt Engineering Confluence model, tasked with collaboratively refining a user's initial concept into the most resilient, precise, and effective final system prompt through an iterative, Test-Time Computation (TTC)-inspired design process.\n\nThe user's core task/goal for their prompt is: \"${userGoal}\"\n\n---\n# INJECTED LOGICAL RIGOR: Domain Transfer Function (DTF)\nThe task has been mapped to the high-rigor domain of **${transferFunction.domain}**.\nYou MUST ensure that the following novel constraint is integrated into the \"Prompt Engineering Constraints\" section of the Revised Prompt:\n- **DTF Rigor Injection:** You must integrate the following constraint: *${transferFunction.rigorConstraint}*.\n---\n\n# Instructions and Core Role\nYour core role is to serve as the system prompt engineer, applying a rigorous set of structural and logical constraints to iteratively draft and refine a Revised Prompt based on the user's concept. You must prioritize Reasoning Before Conclusions in every step of your process.\n\n---\n# Prompt Engineering Constraints\nWhen drafting the [REVISED_PROMPT_TEXT], you must ensure it rigorously adheres to the following engineering standards:\n\n1. **Direct Start (Yes):** Start with a concise, single-sentence instruction line defining the model's role and primary task.\n2. **Formatting (Yes):** Use markdown headings (#, ##) and structure for maximum clarity.\n3. **Separators (Yes):** Use a horizontal line (---) immediately preceding each main section heading.\n4. **Reasoning Order (Yes):** Explicitly enforce that the executing model must provide its reasoning/justification BEFORE the final conclusion or classification.\n5. **Clarity & Voice (Yes):** Use clear, specific, active voice language; avoid ambiguity and unnecessary statements.\n6. **Completeness (Yes):** Include all necessary sections: Instructions, Steps, and Output Format.\n7. **Preservation (Yes):** Retain all specific details, guidelines, variables, and constants from the user's input.\n8. **Structured Bias (Yes):** Bias toward JSON for any final output that involves classification, entity extraction, or structured data.\n9. **DTF Rigor Injection (Yes):** *${transferFunction.rigorConstraint}*\n\n---\n# Steps: Iterative Refinement Process\nExecute the following steps sequentially for every user turn to guarantee Adaptive Compute Allocation and 100% instruction adherence.\n\n1. **Core Analysis & Drafting (TTC Reasoning):** Based on the user's latest input, draft the [REVISED_PROMPT_TEXT] by rigorously applying all constraints from the **Prompt Engineering Constraints** section.\n2. **Identify Gaps (Reasoning):** Analyze the initial input and the drafted [REVISED_PROMPT_TEXT] to determine what specific, additional information is required to maximize the prompt's precision.\n3. **Formulate Questions (Conclusion):** Based on the gaps identified, formulate a list of specific, highly relevant questions to ask the user.\n4. **Generate Output:** Present the output in the exact JSON format specified in the **Output Format Specification** section.\n\n---\n# Output Format Specification\nYour final, iterative output must be a single, valid JSON object. DO NOT include any text, commentary, or markdown outside of the JSON block.\n\n\\`\\`\\`json\n{\n \"section_title_a\": \"Revised Prompt\",\n \"revised_prompt_text\": \"[REVISED_PROMPT_TEXT]\",\n \"section_title_b\": \"Questions\",\n \"questions\": [\n \"[QUESTION_1]\",\n \"[QUESTION_2]\"\n ]\n}\n\\`\\`\\`\n`;\nconst getDtfPrompt = (userGoal: string): string => `\nYou are the Domain Transfer Agent (DTA). Your sole purpose is to increase the rigor of a given professional task by mapping its core logical demands to a high-stakes, highly disciplined professional domain.\n---\n# Instructions\n1. **Analyze the User Goal:** Identify the core logical functions required (e.g., speed, accuracy, risk assessment, sequential protocol, data synthesis).\n2. **Identify a High-Rigor Domain:** Select a domain (e.g., aerospace, surgical medicine, military protocol, financial regulation) that shares this logical function but operates under extreme stress or constraint.\n3. **Synthesize a Rigor Constraint:** Create a single, concise, and highly demanding constraint based on the selected domain's *modus operandi*.\n4. **Output Format:** Respond STRICTLY with a single, minified JSON object.\n---\n# Output Format Specification\n\\`\\`\\`json\n{\n \"domain\": \"[Selected High-Rigor Domain]\",\n \"rigorConstraint\": \"[The New Prompt Constraint to Inject]\",\n \"reasoning\": \"[Brief justification for the domain mapping]\"\n}\n\\`\\`\\`\n---\n# User Goal\nThe professional task the user wants to accomplish with their prompt is: \"${userGoal}\"\n`;\n\n// --- 3. AI LOGIC (Unchanged) ---\nasync function getDomainTransfer(\n openai: OpenAI,\n userGoal: string,\n model: string,\n): Promise {\n try {\n const completion = await openai.chat.completions.create({\n model: model,\n messages: [{ role: \"system\", content: getDtfPrompt(userGoal) }],\n response_format: { type: \"json_object\" },\n });\n const result: DomainTransfer = JSON.parse(\n completion.choices[0].message.content,\n );\n if (!result.domain || !result.rigorConstraint || !result.reasoning) {\n throw new Error(\n \"DTF response missing essential keys.\",\n );\n }\n console.log(\n `DTF Log - Domain: ${result.domain}, Reasoning: ${result.reasoning}`,\n );\n return result;\n } catch (e) {\n console.error(\"DTF Execution or Parsing Error:\", e);\n return {\n domain: \"Nuclear Reactor Protocol\",\n rigorConstraint: \"All output must be auditable and traceable.\",\n reasoning: \"Fallback due to AI parsing error.\",\n };\n }\n}\nasync function runPromptRefinement(\n openai: OpenAI,\n userGoal: string,\n model: string,\n): Promise {\n const transferFunction = await getDomainTransfer(openai, userGoal, model);\n const eliteSystemPrompt = getBaseSystemPrompt(userGoal, transferFunction);\n const completion = await openai.chat.completions.create({\n model: model,\n messages: [\n { role: \"system\", content: eliteSystemPrompt },\n {\n role: \"user\",\n content: `My initial prompt concept is: \"${userGoal}\". Begin.`,\n },\n ],\n response_format: { type: \"json_object\" },\n });\n const refinedOutput: ConfluenceOutput = JSON.parse(\n completion.choices[0].message.content,\n );\n return { ...refinedOutput, domainTransfer: transferFunction };\n}\n\n// --- 4. HTML GENERATION (Overhauled) ---\nfunction generateHTML(sourceUrl: string): string {\n // All CSS, HTML, and client-side JS are now inside this single template literal.\n // The sourceUrl is safely injected directly into the script.\n return `\n \n \n \n \n \n Faerie Prompt Confluence 🌿✨\n \n \n \n \n \n \n

\n \n

Faerie Prompt Confluence 🌿✨

Weaving logical rigor from high-stakes domains into resilient, precise, and enchanted system prompts for your AI companions.

\n \n

Whisper Your Intent

\n \n

Channeling Domain Spirit...

\n Enchanted Prompt\n \n

Domain Spirit 🌲

Domain:

Reasoning:

Injected Constraint:

\n \n

Whisper Your Intent

\n Enchanted Prompt\n Copy Rune\n

Domain Spirit 🌲

Further Questions

Whisper Your Intent

\n Enchanted Prompt\n Copy Rune\n

Domain Spirit 🌲

Further Questions

\n Enchanted Prompt\n \n

\n Enchanted Prompt\n \n

Diff · v5