The TEC Prompt Blueprint
- 1 day ago
- 4 min read

A Fill-In-The-Blank Format That Produces Better AI Output Without Guesswork
Most “bad AI output” starts with a vague prompt.
People type one sentence, hope the tool reads their mind, and then blame the model when the response is generic, overly confident, or unusable.
The fix is simple: stop treating prompts like one-off requests and start using a repeatable structure.
This post gives you the TEC Prompt Blueprint: a fill-in-the-blank prompt format you can reuse for emails, SOPs, proposals, meeting recaps, and internal planning, plus five before/after upgrades you can copy today.
Why It Matters
When a prompt is unclear, AI fills gaps with assumptions. That’s where problems come from:
• Outputs that sound polished but miss the point
• “Helpful” details that were never true
• Wrong tone for the audience
• Extra fluff instead of usable structure
A strong prompt doesn’t just ask for an answer. It defines the job:
• Who it’s for
• What success looks like
• What to include and avoid
• How the output should be formatted
TEC Tip: If you want reliable output, don’t ask AI to “write.” Ask it to follow a specification.
A Quick Real-World Scenario
A team member needs to email a client about a change request. They type:
“Write an email about the change request.”
The draft comes back wordy, over-confident, and it invents a timeline.
They rerun the request using the TEC Prompt Blueprint. The new draft:
• matches the client tone
• includes only confirmed facts
• clearly lists next steps and approvals
• avoids accidental commitments
Same tool. Different prompt.
The TEC Prompt Blueprint (Fill-In-The-Blank)
Copy, paste, and fill this in. The bracketed fields are the only parts you change.
TEC Prompt Blueprint “Help me create a [deliverable].
Audience: [who this is for]
Goal: [what this should accomplish in 1 sentence
]Context: [2–5 bullets of relevant background, no sensitive identifiers]
Must Avoid: • [things that would be risky, inaccurate, off-brand, or inappropriate]
Constraints: • Do not invent facts. If information is missing, write ‘Not specified’ and ask clarifying questions. • Keep commitments conservative (no promises on timelines, pricing, or scope unless provided).
• Keep it professional and on-brand.
Output Format: [bullets / table / email draft / step-by-step / outline]
Length: [short / medium / long]
Tone: [neutral / friendly / firm / executive]
Quality Check: Before finalizing, list any assumptions you made and any details that should be verified.”
TEC Tip: The “Must Include / Must Avoid” section is where prompts become safe and repeatable.
Five Before/After Prompt Upgrades (Copy These)
1) Email Draft (Client Update)
Before: “Write a client update email.”
After: "Help me create a client update email.
Audience: Client sponsor
Goal: Provide status and next steps without overpromising
Context:
• Project is on track overall
• One deliverable is delayed due to [reason provided]
• Next milestone depends on client feedback
Must Include:
• Current status in 3 bullets
• Next steps with owner and date placeholders
• A clear ask for the client
Must Avoid:
• Guarantees on dates
• Blame language
Output Format: Email draft with subject line + 2 tone options
Tone: Professional, calm, confident
Quality Check: List assumptions and items to verify.”
2) SOP Draft (Internal Process)
Before: “Create an SOP for onboarding.”
After: "Help me create an SOP.
Audience: Internal team members
Goal: Standardize onboarding so tasks don’t get missed
Context:
• Roles involved: [IT], [HR], [Manager]
• Tools: [email], [ticketing], [M365], [device management]
Must Include:
• Steps grouped by owner
• Required approvals
• A checklist at the end
Must Avoid:
• Tool-specific instructions we don’t use
• Anything that assumes admin access
Output Format: Numbered steps + checklist
Tone: Clear and direct
Quality Check: Identify any missing inputs.”
3) Proposal Outline (Scope First, Details Later)
Before: “Write a proposal for our services.”
After: "Help me create a proposal outline.
Audience: Prospective client decision-maker
Goal: Define scope and outcomes clearly
Context:
• Client needs: [brief bullets]
• Constraints: [budget/timeline if provided]
Must Include:
• Objectives and success metrics
• Scope and out-of-scope
• Assumptions and dependencies
• Next steps
Must Avoid:
• Pricing or timelines not provided
• Overpromising outcomes
Output Format: Structured outline with headings
Tone: Executive and practical
Quality Check: List assumptions and questions to clarify.”
4) Meeting Recap (Actionable, Not Fluffy)
Before: “Summarize these notes.”
After: “Help me create a meeting recap.
Audience: Internal team
Goal: Capture decisions, actions, and risks clearly
Context: Notes below
Must Include:
• Decisions
• Action items (Owner + Due Date placeholder)
• Risks/blockers
• Next meeting info
Must Avoid:
• Inventing decisions or owners
• Long paragraphs
Output Format: Bullets with clear sections
Tone: Neutral, concise
Quality Check: Flag anything that needs verification.”
5) Troubleshooting Response (Ticket Reply)
Before: “Respond to this ticket.”
After: "Help me create a ticket response.
Audience: End user
Goal: Provide clear steps and set expectations
Context:
• Issue: [summary]
• What we tried: [bullets]
• Environment: [device/app if relevant]
Must Include:
• Step-by-step actions
• When to reply back
• What we’ll do next if it doesn’t work
Must Avoid:
• Blame language
• Unsupported root-cause claims
Output Format: Short email-style reply
Tone: Helpful and calm
Quality Check: Identify assumptions.”
Practical Steps To Make This Stick
Step 1: Make The Blueprint The Default
If your team uses AI weekly, don’t rely on memory. Put the blueprint where work happens:
• A pinned note in Teams/Slack
• A shared doc
• A snippet in your ticketing tool
Step 2: Teach One Habit
When someone says “AI gave me something weird,” ask one question:
• “What did you tell it to do, exactly?”
That quickly shifts the culture from “AI is random” to “prompts are specifications.”
TEC Tip: The goal isn’t perfect prompts. It’s consistent prompts.
How TEC Can Help
• Build role-based prompt blueprints for email, tickets, proposals, SOPs, and leadership updates
• Create safe prompt libraries that reduce improvising and accidental commitments
• Train teams on practical prompting habits that improve output quality fast
• Align prompt workflows with review standards so AI drafts don’t bypass approvals

