thinkhard.SI

Prompt repository Managers & leads

Triage your team's incoming requests against its capacity

Your team has more requests than people, and you need a ranked list plus honest replies to the people who will have to wait.

Registry no.
MGT-020
Collection
Managers & leads
Laid out for
Any chat model
Score
10 out of 10
Blanks to fill
4
Added

Managers & leadsAny chat

The prompt

Score 10 of 10, Excellent

Use it when Your team has more requests than people, and you need a ranked list plus honest replies to the people who will have to wait.

  • {{decision_maker}}
  • {{team_capacity}}
  • {{priorities}}
  • {{requests}}
Full prompt and why it works
Prompt
Help me triage the requests waiting for my team, so that I can agree priorities with {{decision_maker}} and give every requester a straight answer. The team's capacity: {{team_capacity}}.

Our priorities this quarter, in order:
"""
{{priorities}}
"""

The requests, with who asked, what they want, any deadline and my rough estimate of effort:
"""
{{requests}}
"""

Produce:
1. A table with one row per request and the columns Request, Requester, Fits which priority (or "none"), Effort, Deadline, Recommendation (do now, schedule for later, needs a decision, or decline) and a one-line reason.
2. A capacity check: add up the effort of everything marked "do now" and compare it with the team's capacity. If it does not fit, say which items to move and why.
3. Trade-offs for {{decision_maker}}: up to three decisions only they can make, each written as "If we do X, then Y slips."
4. A reply for each requester whose request is not marked "do now", under 80 words each. For later or declined: what will happen and when, or why not, and what they could do instead. For a pending decision: who is deciding and when they will hear.

Rank by fit with the priorities first, then deadline, then effort. The seniority of the requester, or how often they asked, is not a reason on its own; if it seems to matter, list it as a decision for {{decision_maker}}.

The replies should be direct and respectful, with no vague "we'll try to fit it in". Do not promise dates the capacity check does not support. For example, with invented details, a decline that has the right shape: "We can't take this on this quarter because the team is committed to the billing migration. If it can wait until April, I'll put it first in our next planning round."

If a request has no effort estimate or deadline, write "[ESTIMATE NEEDED]" or "[DEADLINE?]" in its row and do not invent one. If two priorities conflict, flag it rather than choosing between them.

Variables

{{decision_maker}}
Who sets priorities with you, for example "my director" or "the operations leadership team".
{{team_capacity}}
People and time available, for example "four analysts, about 30 working days free over the next six weeks".
{{priorities}}
Your team's agreed priorities or goals for this period, most important first.
{{requests}}
Each waiting request: who asked, what they want, any deadline, and your rough effort estimate in days or hours.

Why it works

  • What good looks like. Requests are ranked by fit with the agreed priorities before deadline or effort, so the loudest or most senior requester does not win by default.
  • Length. The capacity check adds up the effort of everything marked do now and compares it with the time the team has, which turns "we're busy" into a number.
  • If unsure. Missing estimates and deadlines are marked rather than invented, and conflicting priorities go to the person who owns them.

How to use it

  1. Press Fill in variables and type your details into each blank. What you type stays in this browser.
  2. Copy the filled prompt into the AI tool you already use, such as ChatGPT, Claude or Gemini, or open it in the Grader or the Prompt Tester here.
  3. Before you rely on it, try it on three or four real examples. The Prompt Tester runs them all at once and shows which ones go wrong.
Use it from code

Get this prompt as JSON from the free REST API, with your own text in the blanks, or let an AI assistant fetch it through the MCP server. No key.

curl -s https://thinkhard.si/api/v1/prompts/team-request-triage

With the MCP server added to Claude, ChatGPT, Cursor or VS Code, call get_prompt with {"prompt": "team-request-triage"}, or just ask for it by name.