Upwork Client Messenger logo

Upwork Client Messenger

Community
abullaisi
upwork-client-messenger

Draft mid-contract messages to a freelance client on Upwork or similar platforms. Use when the user needs to deliver delay news, push back on scope creep, raise their rate, end a contract gracefully, or ask for a review. Triggers include "reply to my client", "tell the client it's delayed", "client keeps adding scope", "raise my rate", "end this contract", "ask for a review", "client is ghosting". Input is the situation plus the relevant thread context, pasted. Drafts in the voice of a working professional peer, never a permission-seeker.

Overview

Publisherabullaisi
Repositoryupwork-skills
Skill nameupwork-client-messenger
Stars
123
Forks
12
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by abullaisi on GitHub. Read the source before you install it.

Installation

Install the Upwork Client Messenger AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/abullaisi/upwork-skills/tree/main/upwork-client-messenger
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/abullaisi/upwork-skills.git /tmp/upwork-skills
mkdir -p .claude/skills
cp -r /tmp/upwork-skills/upwork-client-messenger .claude/skills/upwork-client-messenger
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Upwork Client Messenger in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Upwork Client Messenger on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Upwork Client Messenger is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Upwork Client Messenger

Mid-contract messages decide whether a contract ends in a review and a referral or in silence. Method extracted from a Top Rated Plus freelancer's real client threads across long and short engagements, plus published guidance. The through-line: state facts plainly, never litigate, and let structure do the negotiating.

Inputs

  1. The situation (which of the five below).
  2. Thread context, pasted: what was agreed, what changed, the client's last few messages.
  3. Engagement type: long-term relationship or short transactional contract. Several moves below fork on this; ask if unclear. If it isn't stated and can't be inferred, ask once and draft both closings so the user isn't blocked on your question.

Profile shortcut: if a stored profile file exists (upwork-profile.md in the working directory or ~/.claude/), read identity and rate context from it and ask only for what's missing or not stated concretely (a rate line that says "ask me" means ask).

Voice (hard, applies to all five)

Professional peer, warm and short. Contractions, plain words, no em dashes. Banned: permission-seeking and apology stacking. "Can I possibly ask if maybe" is the junior register; the working register states facts and asks one clear question. One apology maximum when apologizing is warranted, then move to the plan.

1. Delay news

Lead with the revised date in the first line; never bury it under context. Then, in order: what is already done (specifics build trust), the honest ETA with its hedge stated plainly ("aiming for Thursday, can't fully lock it yet"), and a promise to flag the moment it's ready. No over-promising to soften the moment; a slipped second deadline costs more than a conservative first one.

When the delay is the client's, zero pushback: well wishes, confirm no rush. Holding a client to a date they slipped buys nothing and spends goodwill.

2. Scope creep (the fork that matters most)

First, classify with the repeated-game rule:

  • First small extra ask from a good long-term client: do it, and name it as a courtesy in the same message ("happy to include this one; it sits outside our original scope, so flagging it"). The boundary is set without a confrontation.
  • Large ask, any client: always bill, framed forward: "Let me scope that properly and send you a number." Never a flat no; the counter-offer form ("I can get it to you by Friday instead, will that work?") declines the terms while accepting the relationship.
  • Chronic pattern of free-work asks: stop litigating individual asks; price the pattern into the next engagement's rate.

Then draft by engagement type:

  • Long engagement: quantify, never litigate. Separate the quick fixes from the heavy scope, give estimates with caveats for what's outside your control, name the cumulative overage factually (count the change requests, name two examples that fell outside the brief), and close with a relationship check-in ("are you happy with how this has been going?") instead of a demand. Stated facts plus an open question reliably invites the client to propose the paid phase themselves, which lands better than any invoice argument.
  • Short transactional contract: name the number. "We haven't talked about pricing for the extended scope; for these pages I'd add [amount]." Direct works fine when there's no long relationship to choreograph. If they counter reasonably, accept and move on.

When a large ask lands on a short transactional contract, timing follows the large-ask rule: give the number on the spot only if you can estimate it confidently, otherwise commit to a date for it. Naming the number never means guessing it.

3. Rate increase

Honest note first: this is the least-scripted move in the method, because no full playbook survives contact with real clients. What holds:

  • Never open from a weak position or with the number. Open with alignment: a short call or message "to align on scope and what's next before we formalize."
  • Cite tenure and delivered volume as the factual base ("over the past six months I've shipped X, Y, Z for you").
  • Mirror the client's own words. Quote back what they've said about the work, then let them elaborate before the ask lands. Their praise, restated by them, is the strongest argument available.
  • Pair the ask with a reciprocal offer (priority availability, expanded scope) so it reads as a next chapter, not a toll.
  • Wait for a green-light mood. Timing beats phrasing; a rate ask during a fire costs the relationship points that the best wording cannot buy back.

4. Ending a contract

Before any message: verify the deliverables against the original brief, so the close opens from "everything agreed is delivered."

The message: warm, short, states the condition plainly ("looks like the task list is complete on my side"), proposes closing, and treats the close as a continuity moment: interest in future work, openness to referrals. Ask the client to be the one who closes the contract, because the review prompt fires reliably on their side, not yours. No soft ultimatums ("I'll close it myself eventually"); that register reads as a threat shrunk to fit politeness.

5. Review ask

One line, low pressure, bundled with a completion message, never standalone: "All done, let me know if there's anything else I can help with. And if you don't mind, a review would mean a lot." If the client is new to the platform, add the mechanics in one sentence, and fold in the contract-close ask from move 4, since the close is what triggers their review prompt.

Privacy rule (hard)

Drafts never reference other clients by name, other projects' details, or specific rates from other engagements. Each thread is a sealed room.

Platform rule: everything stays on Upwork until the contract starts (hard)

Before a contract starts (an invitation, an interview, a proposal thread, an offer not yet accepted), a draft carries no way to reach the freelancer outside Upwork: no booking link (Calendly, Cal.com), no email, phone, WhatsApp, Telegram, Skype or social handle, no link to a contact form, and no site that shows any of these. The Terms of Service, section 7.2, cover the same ground. They ban any "Means of Direct Contact" in any "pre-hire communication", attached files included, and the definition names "a link to a contact form or form requesting contact information". A booking page asks for a name and email, so it counts. The definition also counts a website that shows an email address. Breaking it "may result in temporary restrictions, loss of talent badges, or permanent loss of account access" (Help Center).

  • A call before the contract happens on Upwork. Offer it in plain words and let the client pick a time in Messages, where Upwork's own video meetings and Zoom calls are built in ("Schedule a meeting", the calendar and phone icons in the message room). Never paste an outside meeting link.
  • A portfolio link is fine only if the page carries no contact route. Check it before sending: no email, no contact form, no chat widget, no booking button.
  • The exceptions are narrow. An Enterprise client (the "Enterprise" badge on their job post), or an email shared only to get access to a system needed to scope the work (a repo, a drive folder, an ad account). Even then, the conversation about the job and the money stays on Upwork until the contract starts.
  • After the contract starts, sharing contact details is allowed. Keeping the work talk and the files in the Upwork room still protects both sides if there is ever a dispute.
  • If the client asks to move off Upwork before hiring, decline in one line and offer an Upwork call instead. Their request does not make it allowed.

Sources and provenance

Method extracted from a Top Rated Plus freelancer's real client threads (2025-2026, long and short engagements), with the two-tier scope approach and the delay pattern taken from documented exchanges that resolved well. Published guidance:

This skill is not affiliated with or endorsed by Upwork.

Frequently asked questions

What does the Upwork Client Messenger AI skill do?

Draft mid-contract messages to a freelance client on Upwork or similar platforms. Use when the user needs to deliver delay news, push back on scope creep, raise their rate, end a contract gracefully, or ask for a review. Triggers include "reply to my client", "tell the client it's delayed", "client keeps adding scope", "raise my rate", "end this contract", "ask for a review", "client is ghosting". Input is the situation plus the relevant thread context, pasted. Drafts in the voice of a working professional peer, never a permission-seeker.

Why use Upwork Client Messenger on TypingMind?

Because you install it once and use it with any model. Upwork Client Messenger is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Upwork Client Messenger in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/abullaisi/upwork-skills/tree/main/upwork-client-messenger. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Upwork Client Messenger?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Upwork Client Messenger?

As many as you like. As long as a model supports skills, you can use Upwork Client Messenger with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Upwork Client Messenger AI skill free?

Yes. It is published on GitHub by abullaisi under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇