# AI Mode ### Quick overview AI Mode lets you interact with your Tability data using natural language prompts. Instead of manually navigating plans, goals, and check-ins, you can ask questions and get quick answers, summaries, and suggested next steps. It is built to help you move faster from data to decisions. *** ### What is AI Mode AI Mode is a prompt-based assistant in Tability that helps you analyse and act on your workspace data. You can use it to: * Summarise progress across plans, goals, and outcomes * Spot risks and blockers from recent check-ins * Draft weekly updates and leadership summaries * Explore trends with follow-up questions Think of AI Mode as a faster, conversational layer on top of your existing Tability workflow. *** ### Beta limitation AI Mode is currently in beta. During beta: * Responses can sometimes be incomplete or imprecise * Results depend on the quality and freshness of your goals and check-ins * Some complex prompts may need follow-up questions for the best output * Features and behavior may evolve quickly Use AI Mode as decision support, and validate critical conclusions before broad sharing. #### 🎁 How to get more prompts during the beta During the beta, AI Mode is free and available to everyone, with a monthly prompt limit. * **Default limit:** 30 prompts per month * **Increased limit:** 150 prompts per month To increase your limit to 150 prompts per month: 1. Go to 2. Publish a 5-star review on G2 3. Send a screenshot of your published review via Tability support chat Once validated, your monthly prompt allowance can be increased. *** ### How to use AI Mode #### Basic workflow 1. Open AI Mode in Tability 2. Ask a specific question (include team, plan, and timeframe when relevant) 3. Review the response 4. Ask follow-up prompts to refine the output 5. Turn insights into action (update goals, assign owners, prepare reviews) #### Prompt tips * Be specific about scope (team, objective, timeframe) * Ask for output format (bullets, table, action plan, summary) * Ask AI Mode to explain why (signals, trends, blockers) *** ### Example prompts #### For managers ``` Summarize this week's progress across all Engineering objectives in 5 bullets. ``` ``` Which KRs are at risk this month, and what are the top blockers? ``` ``` Draft a leadership update for our Q2 plan with wins, risks, and asks. ``` ``` What changed in confidence scores since last week for Product and Design? ``` ``` Recommend 3 focus areas for next week based on current check-ins. ``` #### For individual contributors ``` Summarize my KRs and what needs attention this week. ``` ``` Based on my recent check-ins, what should I prioritize next? ``` ``` Draft a weekly update I can post in Slack from my current progress. ``` ``` What blockers have I mentioned in the last 2 check-ins? ``` ``` Help me rewrite my KRs to be more measurable and specific. ``` *** ### Advanced prompt examples #### Exec-ready retrospective ``` Create an executive retrospective for [Plan Name] Use only available workspace data (plans, objectives, key results, check-ins, metrics, projects, dependencies, decisions). If evidence is missing, write: Data not available. Formatting rules (strict): - Use Markdown headings. - Do not produce long bullet lists. - Max 3 bullets per section (except checklists: max 6). - Prefer: 1 short framing paragraph, then compact table, then up to 3 bullets. - Keep paragraphs to 2–4 lines. - Use labels: ✅ On track, ⚠️ Risk, ❌ Off track, 📌 Decision. - End with “Reader TL;DR” (max 3 bullets). Traceability rules (strict): - Every non-trivial claim must be justified. - Link real workspace items whenever possible - If no source exists, write: Unverified (no linked item). Output sections in this exact order: ## Executive Summary Short paragraph + up to 3 bullets with citations. ## KRs summary Table: KR | Target | Actual | Status | Commentary ## KR Missed or Partially Achieved Short paragraph + table: KR | Gap | Root Cause | Recovery Option ## Key Risks and Root Causes Table: Risk | Symptom | Root Cause | Owner | Next Mitigation ## Cross-Team Dependencies Table: Dependency | Helped/Slowed | Team | Timeline Effect ## Customer / Business Impact Short paragraph + up to 3 evidence-based bullets. ## Top 5 Lessons Learned Numbered list of exactly 5 lessons, one sentence each. ## Recommended Priorities for Next Quarter Table: Priority | Why now | Expected Impact | Owner ## Executive Decisions Needed Up to 3 bullets, each starting with 📌. ## Reader TL;DR Up to 3 bullets. ``` #### Parent/dependent plan misalignment ``` Assess alignment between parent plan [Parent Plan Name] and dependent plan [Dependent Plan Name]. Use only available workspace data. If evidence is missing, write: Data not available. Formatting rules (strict): - Use Markdown headings. - Avoid long bullet lists. - Max 3 bullets per section. - Prefer short paragraph + compact table + brief takeaways. - Keep paragraphs to 2–4 lines. - Use labels: ✅ On track, ⚠️ Risk, ❌ Off track, 📌 Decision. - End with “Reader TL;DR” (max 3 bullets). Traceability rules (strict): - Every non-trivial claim must be justified. - Link real workspace items whenever possible - If no source exists, write: Unverified (no linked item). ## Alignment Verdict Give misalignment score: Low | Medium | High, plus 3–4 lines of justification. ## Scorecard by Dimension Table: Dimension | Status | Evidence | Severity (1-5) Include these dimensions: - Outcome alignment - Ownership coverage - KPI consistency - Sequencing health ## Key Issues Table: Issue | Dimension | Owner | Recommended Fix | ETA ## 30-Day Realignment Plan Week-by-week table (Week 1–4): Week | Actions | Owner | Success Signal ## Escalations Needed Up to 3 bullets starting with 📌. ## Data Gaps Only if needed; max 3 bullets. ## Reader TL;DR Up to 3 bullets. ``` #### “Win the Week” coach mode ``` Act as my weekly performance coach using my active objectives and latest check-ins. Use only available workspace signals. If signals conflict or are incomplete, state assumptions explicitly and write Data not available where needed. Formatting rules (strict): - Use Markdown headings. - No long bullet walls. - Max 3 bullets per section (checklist max 6). - Prefer short paragraph + compact table. - Keep paragraphs to 2–4 lines. - Use labels: ✅ On track, ⚠️ Risk, ❌ Off track. - End with “Reader TL;DR” (max 3 bullets). - Keep total output under 350 words. Traceability rules (strict): - Every non-trivial claim must be justified. - Link real workspace items whenever possible - If no source exists, write: Unverified (no linked item). ## Weekly Focus One short paragraph naming the week’s theme. ## Top 3 Priorities Table: Priority | Why this matters now | Time Block | Success Signal ## Quick Win (<2h) and High-Leverage Risk Reducer Table: Type | Task | Effort | Expected Payoff ## Draft Status Update Provide a ready-to-send update in 4–6 sentences, no bullets. ## Friday Self-Review Checklist Checklist with max 6 items. ## Anti-Pattern Warning One short paragraph: what is likely to derail the week and what to do instead. ## Reader TL;DR Up to 3 bullets. ``` ### Best practices * Keep prompts clear and scoped * Include context like team and plan name * Use follow-up prompts to sharpen the result * Reuse prompts that work well for recurring reviews * Verify key decisions with source data when stakes are high --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://guides.tability.io/docs/become-a-tability-power-user/features/ai-features/ai-mode.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.