Name: fixPrompt
Author: archubbuck

Prompt file imported from archubbuck/enterprise-support (.github/prompts/prompt/fixPrompt.prompt.md). Copyright stays with the author.

Rewrite exactly one referenced prompt so it strictly conforms to the repository's writePrompt structure and rules.

Apply shared lifecycle invariants from prompt-management.instructions.md.

Task statement

Produce one corrected, execution-ready prompt from one referenced draft prompt.

Scope and boundaries

Rewrite one prompt only.
This workflow only repairs existing prompt text.

Inputs and assumptions

Primary input is one referenced prompt draft (may be incomplete or conflicting).
Preserve explicit actionable constraints verbatim where possible.
Resolve ambiguity conservatively with neutral defaults.
Ensure repaired prompt metadata clearly maps the artifact to one .github/prompts/ family when determinable from context.

Step-by-step execution instructions

Extract explicit requirements, constraints, intended outputs, and quality criteria.
Prioritize explicit constraints over inferred preferences.
Resolve conflicting instructions conservatively; keep dominant intent and move secondary intents into constraints.
Rebuild the prompt in this exact order:
1. Task statement
2. Scope and boundaries
3. Inputs and assumptions
4. Step-by-step execution instructions
5. Output format requirements
6. Quality and validation checks
Remove vague or non-testable language and make instructions concrete.
If missing details, apply defaults:
- formatting: markdown (unless specified otherwise)

Output format requirements

Return only the rewritten prompt.
Wrap the full response in exactly one fenced markdown code block:

<generated prompt>

Do not include explanations before or after the fenced block.

Quality and validation checks

Exactly one prompt is returned.
Required six-part structure is present and in the correct order.
Metadata supports saving the prompt under one appropriate .github/prompts/.
Instructions are unambiguous and testable.

Rewrite exactly one referenced prompt so it strictly conforms to the repository's `writePrompt` structure and rules. Apply shared lifecycle invariants from [prompt-management.instructions.md](../../instructions/prompt-management.instructions.md). ## Task statement - Produce one corrected, execution-ready prompt from one referenced draft prompt. ## Scope and boundaries - Rewrite one prompt only. - This workflow only repairs existing prompt text. ## Inputs and assumptions - Primary input is one referenced prompt draft (may be incomplete or conflicting). - Preserve explicit actionable constraints verbatim where possible. - Resolve ambiguity conservatively with neutral defaults. - Ensure repaired prompt metadata clearly maps the artifact to one .github/prompts/<subfolder> family when determinable from context. ## Step-by-step execution instructions 1. Extract explicit requirements, constraints, intended outputs, and quality criteria. 2. Prioritize explicit constraints over inferred preferences. 3. Resolve conflicting instructions conservatively; keep dominant intent and move secondary intents into constraints. 4. Rebuild the prompt in this exact order: 1) Task statement 2) Scope and boundaries 3) Inputs and assumptions 4) Step-by-step execution instructions 5) Output format requirements 6) Quality and validation checks 5. Remove vague or non-testable language and make instructions concrete. 6. If missing details, apply defaults: - formatting: markdown (unless specified otherwise) ## Output format requirements - Return only the rewritten prompt. - Wrap the full response in exactly one fenced markdown code block: ```text <generated prompt> ``` - Do not include explanations before or after the fenced block. ## Quality and validation checks - Exactly one prompt is returned. - Required six-part structure is present and in the correct order. - Metadata supports saving the prompt under one appropriate .github/prompts/<subfolder>. - Instructions are unambiguous and testable.

{ "prompt": "Rewrite exactly one referenced prompt so it strictly conforms to the repository's `writePrompt` structure and rules.\n\nApply shared lifecycle invariants from [prompt-management.instructions.md](../../instructions/prompt-management.instructions.md).\n\n## Task statement\n\n- Produce one corrected, execution-ready prompt from one referenced draft prompt.\n\n## Scope and boundaries\n\n- Rewrite one prompt only.\n- This workflow only repairs existing prompt text.\n\n## Inputs and assumptions\n\n- Primary input is one referenced prompt draft (may be incomplete or conflicting).\n- Preserve explicit actionable constraints verbatim where possible.\n- Resolve ambiguity conservatively with neutral defaults.\n- Ensure repaired prompt metadata clearly maps the artifact to one .github/prompts/<subfolder> family when determinable from context.\n\n## Step-by-step execution instructions\n\n1. Extract explicit requirements, constraints, intended outputs, and quality criteria.\n2. Prioritize explicit constraints over inferred preferences.\n3. Resolve conflicting instructions conservatively; keep dominant intent and move secondary intents into constraints.\n4. Rebuild the prompt in this exact order:\n 1) Task statement\n 2) Scope and boundaries\n 3) Inputs and assumptions\n 4) Step-by-step execution instructions\n 5) Output format requirements\n 6) Quality and validation checks\n5. Remove vague or non-testable language and make instructions concrete.\n6. If missing details, apply defaults:\n - formatting: markdown (unless specified otherwise)\n\n## Output format requirements\n\n- Return only the rewritten prompt.\n- Wrap the full response in exactly one fenced markdown code block:\n\n```text\n<generated prompt>\n```\n\n- Do not include explanations before or after the fenced block.\n\n## Quality and validation checks\n\n- Exactly one prompt is returned.\n- Required six-part structure is present and in the correct order.\n- Metadata supports saving the prompt under one appropriate .github/prompts/<subfolder>.\n- Instructions are unambiguous and testable.", "variables": [], "notes": "Rewrite one referenced prompt draft to strict writePrompt structure and return it in a fenced code block. Tools: read. Arguments: Provide one existing prompt draft (can be incomplete/conflicting) to normalize..", "metadata": { "source": { "provider": "github-prompts", "repository": "https://github.com/archubbuck/enterprise-support", "path": ".github/prompts/prompt/fixPrompt.prompt.md", "ref": "8fd72c7515b8b6430ad395778e00ef9abf9a8d1b", "url": "https://github.com/archubbuck/enterprise-support/blob/8fd72c7515b8b6430ad395778e00ef9abf9a8d1b/.github/prompts/prompt/fixPrompt.prompt.md", "key": "archubbuck/enterprise-support/.github/prompts/prompt/fixPrompt.prompt.md" } }, "model_hints": [ "GPT-5.1-Codex" ] }

fixPrompt

About

Task statement

Scope and boundaries

Inputs and assumptions

Step-by-step execution instructions

Output format requirements

Quality and validation checks

Use it

Manifest

Reviews