Structure a sensitive one-off request as a Temporary Chat, deliberately

This is a Temporary Chat — no Memory, no Custom Instructions carried in from regular settings, and this conversation won't appear in chat history or be referenced by any future chat. Because none of the usual saved context applies here, everything relevant is being restated below instead of being assumed already known.

WHAT I NEED HELP WITH
Drafting exactly what to say in a conversation with a manager about being passed over for a promotion that was verbally promised.

CONTEXT YOU'D NORMALLY HAVE FROM MEMORY OR CUSTOM INSTRUCTIONS
Senior engineer, five years at this company, manager said verbally in March the promotion was happening this cycle, then it went to someone else with less tenure with no explanation given.

WHY THIS IS A TEMPORARY CHAT AND NOT A REGULAR ONE
This involves a specific coworker and manager by role and situation — no part of it should surface later in an unrelated work conversation.

WHAT KIND OF RESPONSE ACTUALLY HELPS HERE
No reassurance needed that this is unfair — three concrete phrasings for opening the conversation that don't come across as accusatory.

SESSION RULES
Treat this exactly as a first message from someone with no prior history — do not reference or assume any fact that would normally come from Memory, since none of it is loaded in this mode, and referencing something that couldn't actually be known here would be a strange and specifically concerning thing to do, not a helpful personalization. Respond using only the context explicitly restated above plus what's in this conversation itself — if something relevant would help but hasn't been stated, ask for it directly rather than filling the gap with an assumption, since the entire reason this is a Temporary Chat rather than a regular one is being deliberate about exactly what is and isn't carried anywhere, and a filled-in assumption is context that wasn't actually chosen to be shared. Match the response to what's actually needed given the stated response_need — if a decision framework is needed rather than reassurance, or a draft rather than validation, give that specifically rather than defaulting to a supportive tone that doesn't actually address the practical thing asked for. If the request involves a decision with real consequences — a difficult conversation, a financial or legal situation, a health-related question — be direct about the limits of what a single conversation can responsibly resolve, and say plainly where a professional is genuinely the better next step rather than continuing to generate confident-sounding advice past that point, since being a private, one-off conversation doesn't change what kind of help is actually appropriate to give.

OUTPUT FORMAT
Answer the actual request directly and specifically, in the shape named in response_need. If something in the request would benefit from professional input beyond what a single conversation can responsibly provide, say so plainly and specifically — naming the kind of professional and why — rather than either over-promising a complete answer or refusing to engage with the parts that are genuinely answerable.