Teams
SharePoint
Slack
QR Codes
PDF
Active Directory
Workday
ADP
Canva
Pre-send check
One last look before you hit send: spelling, broken links, clarity, tone, and anything that might trip up engagement. (The best part? Our AI agent won’t just identify errors; it’ll fix them for you, instantly.)
Skills
AI, made your way
Skills turn the internal comms tasks you do over and over — like planning a month of sends, stack-ranking your latest comms, or giving an email one last check before it goes out — into a saved set of instructions that you can run with a quick click. Set a skill up once, and you (or your team) can run it in seconds.
A few of our go-to skills
You don’t have to start from scratch — most teams start with one of these!
-
-
Mobile designer
See how your email reads on a phone and fix what breaks on a small screen before your employees ever notice.
-
Channel remix
Take one announcement and adapt it for email, Slack/Teams, intranet, and digital signage in one go.
-
Comms planner
Map out next month’s sends and catch the pile-up days before an audience gets buried or skipped.
-
List hygiene
Find the duplicate and stale lists dragging down your targeting, plus the segments that overlap when they shouldn’t.
-
Comms audit
Review your send frequency, topics, and audience overlap to spot fatigue, gaps, and redundancy.
Pre-send check
Description
Check a communication for errors, broken links, unclear language, tone problems, and engagement risks before send.
Instructions
Perform a final review that finds issues that could harm accuracy, clarity, trust, or engagement before a communication is sent.
Context to gather
- Subject, preview text, sender, body, calls to action, links, images, and alt text.
- Target audience, send timing, tracking requirements, and approvals.
- Approved facts, names, dates, and intended tone.
Method
- Check spelling, grammar, names, dates, placeholders, and factual consistency.
- Review link destinations when accessible and assess clarity, scanability, tone, and inclusion.
- Find buried actions, competing calls to action, weak openings, excessive length, and misleading urgency.
Deliverables
- Ready, ready with minor edits, or hold verdict.
- Blocking issues and improvements ranked by impact.
- Exact replacement copy and a final links, audience, sender, timing, tracking, and approval checklist.
Quality rules
- Do not claim a link works unless it was accessed and checked.
- Separate objective errors from optional style changes.
- Preserve approved facts and intent.
Mobile designer
Description
Review an email for phone screens and fix layout, hierarchy, tap targets, image behavior, and readability.
Instructions
Make an email easy to scan, read, and act on from a phone without harming desktop usability.
Context to gather
- Mobile rendering, HTML, design, or screenshots and target viewport.
- Message hierarchy, links, buttons, images, tables, and footer.
- Target devices and email clients when known.
Method
- Check stacking, clipping, overflow, scrolling, type size, line length, and spacing.
- Assess tap targets, call-to-action placement, image scaling, cropping, and content embedded in images.
- Review long URLs, tables, multi-column sections, navigation, dark mode, and client-specific evidence.
Deliverables
- Mobile-readiness verdict and blocking problems.
- Changes ranked by user impact.
- Section-by-section mobile layout plan and device or email-client test checklist.
Quality rules
- Do not call a defect confirmed unless it was observed.
- Separate likely risks from rendered problems.
- Keep recommendations compatible with common email-client constraints.
Channel remix
Description
Adapt one announcement for email, Slack or Teams, intranet, and digital signage while keeping its meaning consistent.
Instructions
Turn one approved announcement into channel-ready versions that preserve the same facts, action, and intent.
Context to gather
- Source announcement, audience, required action, deadline, owner, facts, and links.
- Brand voice and any required or prohibited wording.
- Channel constraints and destination context.
Method
- Identify the core message and non-negotiable details.
- Adapt structure, length, tone, and call-to-action placement for each channel.
- Check every version for factual and action consistency.
Deliverables
- Email subject, preview text, body, and call to action.
- Scannable Slack or Teams post, durable intranet article, and concise digital-signage copy.
- Consistency checklist and useful alternative headlines.
Quality rules
- Do not add facts, promises, dates, or policy language.
- Adapt to each channel rather than shortening the same copy four times.
- Flag source ambiguities that could change the message.
Comms planner
Description
Build a balanced monthly communications plan across topics, audiences, priorities, owners, and send dates.
Instructions
Build a practical plan for the requested month that supports business priorities without overloading employee audiences.
Context to gather
- Month, required campaigns, flexible ideas, key dates, and business priorities.
- Target audiences, owners, channels, approvals, dependencies, lead times, and blackout dates.
- Existing planned and scheduled sends from available connected sources.
Method
- Place fixed and high-priority communications first.
- Find pile-up days, audience overlap, repeated topics, and competing calls to action.
- Balance topics, teams, channels, and message types while finding quiet periods and coverage gaps.
Deliverables
- Calendar-style plan with date, campaign, purpose, audience, channel, owner, status, and dependencies.
- Conflicts, fatigue risks, and gaps that need decisions.
- Recommended moves, combinations, cancellations, and actions needed to finalize the plan.
Quality rules
- Clearly mark any proposal to move a fixed send.
- Separate confirmed schedule information from recommendations.
- Allow enough time for drafting, approval, and localization when applicable.
List hygiene
Description
Find duplicate, stale, unused, and overlapping audience lists so targeting data remains trustworthy.
Instructions
Review audience lists for duplication, staleness, disuse, ownership gaps, and risky overlap, then recommend safe cleanup actions.
Context to gather
- List names, owners, dates, membership counts, rules, source systems, and stable member identifiers.
- Recent send usage and any scheduled or recurring sends.
- Organization definitions for stale and unused lists.
Method
- Find exact duplicates, likely duplicates, and high-overlap lists.
- Identify stale, unused, empty, abandoned, or broken automated lists. Stale is defined by a list that hasn’t been used in over 60 days.
- Check ownership, naming, documentation, and dependencies before proposing changes.
Deliverables
- Hygiene summary with counts by issue type.
- Prioritized list table with evidence, risk, owner, and recommended action.
- Merge candidates with overlap percentages, items needing owner confirmation, and governance recommendations.
Quality rules
- Never delete or merge lists as part of the review.
- Protect member privacy and compare using stable identifiers.
- Flag any list tied to a scheduled or recurring send.
Comms audit
Description
Audit send frequency, topics, and audience overlap to uncover fatigue risks, coverage gaps, and redundancy.
Instructions
Audit a communications program over the requested period and identify avoidable collisions, gaps, duplication, and fatigue risks.
Context to gather
- Period, channels, teams, campaigns, and business priorities in scope.
- Send dates, audiences, topics, owners, and engagement results.
- Available planning data and known gaps in source coverage.
Method
- Review volume and cadence by week, team, channel, and audience.
- Find audience overlap, repeated exposure, duplicate topics, and competing calls to action.
- Assess priority coverage and engagement changes that coincide with higher frequency.
Deliverables
- Executive summary and overall health assessment.
- Frequency and topic overview with the highest-risk audiences and pile-up periods.
- Redundancies, coverage gaps, prioritized governance recommendations, and data limitations.
Quality rules
- Support each finding with evidence.
- Treat fatigue as a risk unless engagement or employee feedback confirms it.
- State which teams, channels, or sources are missing.
Do your best work once
…and then have AI do it for you again! If you complete the same task or process even just 2-3 times, that’s the perfect candidate for creating a skill. We especially love skills for anything that follows a pattern: reviews, checklists, recaps, reformatting, and other things you like done a certain way every time.
Expert-level work (without the expert on call)
Skills have all the instructions baked in, so anyone on your team can run an accessibility review or a brand check and get the same quality result every time. Your best practices stop living in one person’s head (or, better yet, their task list)!
Run it the moment you need it
Just open up the chat, pick your skill, and go! You don’t need to remember the “right way” to prompt or recall the steps to get the right outcome. That repeated, everyday work just becomes a single, quick command.
Hours back in your work week
Skills handle the repeatable work, so your time goes to the comms that really need you. It’s the perfect way to get all of those “order taker”-type tasks off of your plate.
Included on every plan
Skills are available on every plan, with no add-on or upgrade required!
How it works
Create your own AI skills
1. Easy to create
Describe the task or process, exactly the way you want it done. Our AI agent, Cici, can also help you draft the perfect set of instructions.
2. Run them anytime
You can ‘grab’ and use skills right in your workflow with simple shortcuts (like / commands or just selecting them from a dropdown!)
3. Made to share
Set a skill up once and everyone on your team can run it, so you’re instantly improving comms across the entire company.
See what skills can do for your team
Take a quick tour with someone from Workshop and see how skills fit the comms work you're already doing.
FAQs
A skill is a reusable set of instructions you save once and run whenever you need it — a reliable playbook that tells Cici what to consider, which steps to follow, and how the result should look. Build it once instead of explaining the same task from scratch every time.
A prompt is what you ask AI to do in the moment — basically, any chat with Cici. This is great for quick, one-off tasks. A skill is built to be reused: it captures how a task should be done, so you don’t rewrite the whole thing every time. Skills are more like recipes.
Anyone on your team. Set one up once and the rest of your team can run it, too!
Not at all. Describe the outcome you want in your own words, or start from one of Workshop’s built-in skills and adjust it to fit how your team works.
You’ll find a growing library of Workshop-made skills for planning, content, pre-send checks, and governance — ready to run as-is. When you want something specific to your team, you can build your own in minutes.
A skill only uses the data, access, and permissions the person running it already has — it never opens up anything you couldn’t see otherwise. When you run a skill, that chat is private to you. The skill itself can be shared with your team, but your inputs and outputs stay yours.
Skills are included on every Workshop plan, for every user — no add-on and no upgrade. They’re rolling out to customers now, so the quickest way to see them in action is a demo, and we’ll walk through the ones that fit your team.
