Astro Prompt Cheat Sheet

Use this page as a practical prompt library for Astro. It is grounded in the seeded demo tenant so the examples use real departments, teams, open roles, and people instead of placeholders.

Astro is tenant-wide — your conversation persists as you navigate between baseline, scenarios, and snapshots. Astro can read from any context and write to active scenarios only.

Astro is strongest when you:

• ask one clear question or one governed change at a time
• review the preview and confirm each proposed change deliberately
• reference specific contexts when asking cross-context questions (e.g. "compare the Growth Team in baseline versus the Q2 restructure scenario")
• use Astro to prepare promotion decisions, then use the normal Promote flow for the final baseline change

Astro now auto-routes complex read prompts to a higher-quality model policy and pre-resolves likely entities server-side before answering. You can help this work even better by naming teams, departments, and scenarios explicitly (including short aliases like Eng for Engineering). Astro also retries failed lookups automatically with alias and scope expansion before it asks you to clarify, so typo-tolerant prompts like Enginering dept or SRE team are now supported more reliably. For workflow and troubleshooting questions, Astro now grounds answers in indexed Orgonaut docs. When the question also depends on the current page or tenant state, Astro will split the answer into Documented guidance and Current tenant data.

Pick the right prompt shape

Use one of these four prompt shapes depending on what you need back from Astro:

• Direct question when you want one grounded answer or comparison, such as staffing, cost concentration, velocity, or scenario status.
• Read-only plan when you want a multi-phase recommendation for an AI transformation, restructure, or broader operating-model shift. This is the right place to get Option A/B/C playbooks, trade-off tables, and a recommended sequence before any change drafting starts.
• Phased scenario program when you are already inside a mutable scenario and want Astro to break a broad restructure objective into sequential governed checkpoints.
• Bundled proposal draft when you already know the coordinated scenario changes that should be reviewed together and confirmed as one bundle.

Astro routes direct questions through its lower-latency lookup path. Read-only plans and bundled proposal prompts use a higher-quality analysis path, so those broader prompts can take a little longer but return richer evidence, playbooks, and execution-handoff guidance. If Astro gives you a restructure plan first, the safest next step is usually to send the recommended apply prompt from that reply instead of rewriting the bundle request from scratch.

Demo reference points

These prompts assume the current demo tenant baseline includes:

• 20 teams and 12 departments
• 115 human actors and 50 agent actors
• an Engineering division split into Platform Engineering, Solutions Engineering, and Product Engineering
• four useful starter teams for AI-first planning: API Platform Team, Developer Experience Team, Growth Team, and Mobile Team
• current open roles in teams such as:
  • API Platform Team -> Software Engineer
  • Developer Experience Team -> Software Engineer
  • Growth Team -> Product Manager
  • Mobile Team -> QA Engineer
  • Infrastructure Team -> Platform Engineer
• current example people you can ask about:
  • Ben Afflake in API Platform Team
  • Jessica Chestane in Developer Experience Team
  • Robert Downer Jr. in Growth Team
  • David Harboard in Mobile Team

The seeded engineering data also makes it easy to compare teams that already have agent support. In the current demo baseline, API Platform Team, Developer Experience Team, and Growth Team each already include three agents, while Mobile Team includes two.

1. Explore the live baseline

Use these prompts in Live when you want a fast picture of the current organisation.

Give me a summary of the live organisation as of today.

Summarise Engineering, including headcount, FTE, monthly cost, and sprint performance.

Compare Platform Engineering, Solutions Engineering, and Product Engineering.

Compare API Platform Team, Developer Experience Team, Growth Team, and Mobile Team on people, agents, open roles, cost, and velocity.

Which engineering teams already have the highest agent mix?

Show me the open roles across API Platform Team, Developer Experience Team, Growth Team, Mobile Team, and Infrastructure Team.

Where does Ben Afflake sit today, and what else is he allocated to?

Show me the current team context for Jessica Chestane, Robert Downer Jr., and David Harboard.

2. Read across contexts

Astro can read from baseline, any scenario, and any snapshot in a single conversation. Use these prompts to compare or pull data from different contexts without switching pages.

Compare the Growth Team in baseline versus the Q2 restructure scenario.

How has Product Engineering changed between the March snapshot and the current baseline?

Show me the headcount delta for API Platform Team across all active scenarios.

What open roles exist in the AI-first SDLC transition scenario that are not in baseline?

Summarise the cost difference for Engineering between the current baseline and the latest snapshot.

3. Ask for product guidance and troubleshooting

Use these prompts when you want grounded product-help guidance rather than a live metric read.

How do I promote a scenario to baseline safely?

How do I fix the latest CSV import errors for this tenant?

How do I connect BambooHR live sync?

Where is the API reference for actors and org units?

Which docs explain Matrix exports and filters?

4. Get insight before you change anything

Use these prompts to narrow the best starting points for an AI-first SDLC transition.

Which engineering teams are the best candidates for an AI-first SDLC transition, based on current agent mix, open roles, and sprint velocity?

If I want quick wins, should I start with API Platform Team, Developer Experience Team, Growth Team, or Mobile Team? Explain why.

Which open roles look like the cleanest first candidates to convert into agent-oriented work?

What risks would you flag before shifting more delivery work in Mobile Team toward agents?

How much of Engineering's current cost sits in Platform Engineering versus Product Engineering versus Solutions Engineering?

Which teams look expensive relative to their sprint output?

What would a cautious AI-first rollout look like for Product Engineering without destabilising delivery?

Build a read-only restructure program first

Start here before you ask Astro to draft any governed proposal bundles.

Astro now returns three built-in strategy playbooks in this flow:

• Option A — augmentation-first
• Option B — automation-pod
• Option C — selective role redesign

Each option includes a trade-off table plus a recommended step sequence with tool references and source citations. Astro also returns an AI-assessed recommended option with a short reason so you can decide whether to proceed with that playbook.

Draft a read-only AI-generated multi-phase restructure plan to make Engineering AI-first over two quarters. Include assumptions, constraints, estimated headcount/FTE/cost deltas, risks, dependencies, and the governed execution handoff.

In the Q2 restructure scenario, build a phased restructure plan that keeps monthly cost flat while increasing platform delivery capacity.

Draft Option A/B/C for an AI transformation in this scenario and recommend one sequence with trade-off tables and tool-cited evidence.

Use the recommended playbook from that plan and draft governed proposal bundles for this scenario. Keep explicit confirmation before execution.

That prompt is the normal handoff from planning into governed execution: Astro keeps the selected playbook context, expands it into bundle previews, and still waits for explicit confirmation before any scenario write happens.

Turn a broad scenario objective into a phased program

Use this when you are already inside a mutable scenario and want Astro to decompose the work instead of asking you for low-level record ids up front.

Astro will draft the phased checkpoint first. Confirming that root checkpoint drafts the first executable governed bundle. After you confirm and execute a phase, Astro can queue the next checkpoint automatically.

In this scenario, reduce management layers in Engineering while keeping monthly cost flat. Draft the phased governed program and stop at the first checkpoint.

In this scenario, flatten Product Engineering by one management layer without reducing platform delivery capacity. Show phase metrics, dependency checks, and rollback points.

In this scenario, realign Engineering for an AI-first operating model over phased governed checkpoints. Keep every write confirmation-gated.

5. Create the planning scenario

Use these prompts to have Astro draft and update a scenario. The flow is governed and review-oriented.

Live baseline and snapshots stay on hard rails. If you want Astro to plan or stage governed changes, start by creating or opening a mutable scenario.

Create a new scenario.

Create a new scenario draft called AI-first SDLC transition for Product Engineering.

Create a new Astro planning scenario called AI-first SDLC transition for Product Engineering. Keep every execution step confirmation-gated.

Update this scenario description to focus on API Platform Team, Developer Experience Team, Growth Team, and Mobile Team, with an effective date of 2026-04-01.

Create an initiative in this scenario called AI-first SDLC enablement.

Summarise the scope and intended outcome of this scenario in plain English.

6. Stage sensible AI-first changes inside the scenario

Astro is most reliable here when each prompt describes one governed change. Start by using open roles and additive agent capacity before you remove or replace human capacity.

Use governed bulk proposals when one intent spans many records

Astro can now expand supported bulk requests into one reviewable proposal bundle instead of making you prompt record-by-record.

When you are in a mutable scenario, Astro will show that it is operating in the Scenario sandbox. That means it can plan broadly and draft supported scenario-scoped changes there, but the confirmation requirement does not change.

For broader restructure goals, Astro can also return a phased scenario program instead of forcing you into one giant bundle or a low-level clarification loop. That is the right path for goals like de-layering, org realignment, or AI-first operating-model shifts.

In this scenario, increase all salaries in Engineering by 10% from 2026-07-01. Show the governed preview before you confirm anything.

In this scenario, redistribute all Scrum Masters evenly across Alpha, Beta, and Gamma. Show before-and-after team coverage and wait for explicit confirmation.

Add agent capacity in the right places

In this scenario, create a new agent actor called Build Copilot 1.

In this scenario, place Build Copilot 1 in Developer Experience Team at 1.0 FTE effective 2026-04-01.

In this scenario, add a USD 2500 cost event for Build Copilot 1 on 2026-04-01 for build and test automation.

In this scenario, create a new agent actor called API Test Agent 1.

In this scenario, place API Test Agent 1 in API Platform Team at 1.0 FTE effective 2026-04-01.

In this scenario, add a USD 3000 cost event for API Test Agent 1 on 2026-04-01 for API contract and regression automation.

Convert open roles before you cut existing people

In this scenario, update the open Software Engineer position in API Platform Team so it becomes an agent-focused platform automation role.

In this scenario, update the open Software Engineer position in Developer Experience Team so it becomes an agent-focused developer tooling role.

In this scenario, update the open QA Engineer position in Mobile Team so it becomes an agent-focused mobile regression testing role.

In this scenario, update the open Product Manager position in Growth Team so it becomes an agent-assisted experimentation operations role.

Ask Astro to propose role replacement cautiously

Use prompts like these when you want Astro to help you stage a replacement, but still keep the conversation explicit and reviewable:

In this scenario, show me a cautious proposal to reduce one Software Engineer role in Growth Team and replace it with agent capacity. Do not confirm anything yet.

Draft a bundled proposal for this scenario: create two platform engineers (Alex Chen and Priya Shah) and move one open QA role from Growth Team to Platform Team. Show all bundle items before confirmation.

In this scenario, show me the staged changes needed to make API Platform Team more AI-first without changing its tech lead coverage.

In this scenario, suggest the lowest-risk human-to-agent changes across API Platform Team, Developer Experience Team, Growth Team, and Mobile Team.

In this scenario, tell me which roles should remain human-led even if we add more agents to Product Engineering.

7. Review the delta properly

Once you have staged a first wave of changes, use Astro to inspect the scenario delta before you ask anyone else to review it.

Summarise the delta for this scenario versus live baseline.

Compare API Platform Team, Developer Experience Team, Growth Team, and Mobile Team before and after this scenario.

What changed in cost, FTE, and delivery capacity across Product Engineering?

Show every staged change in this scenario, grouped by team and by operation.

What are the biggest risks, assumptions, and missing follow-up tasks in this scenario?

Write a stakeholder-ready summary of this scenario for the VP of Engineering.

8. Manage tenant-level settings

Astro can manage tenant-level settings directly because these are not scope-bound. These changes apply tenant-wide regardless of which page you are on.

Add a new department type called Research Lab.

Show me all current team types.

Create a new role type called AI Operations Lead.

Add a new currency: Japanese Yen (JPY).

Update the AI cost for GPT-4o to USD 0.0025 per 1K input tokens.

Create a new sprint called Sprint 24 starting 2026-04-01 and ending 2026-04-14.

List all department types and show which ones are currently in use.

9. Prepare for review and promotion

Astro can help you review and submit the scenario, but it does not approve or promote scenarios to baseline for you.

Prepare a promotion briefing for this scenario: goals, changes, cost delta, capacity delta, and key risks.

What should I verify manually before I promote this scenario to baseline?

List anything in this scenario that looks unsafe, incomplete, or inconsistent.

Submit this scenario for review.

Draft a short note I can paste into the promotion review describing why this scenario is worth promoting.

Prompting patterns that work well

• Name the exact team, department, actor, or scenario you mean.
• Ask Astro to compare a small set of teams rather than all of Engineering at once.
• In scenario mode, use one change per turn when you want reliable previews and confirmations.
• Ask for a bundled proposal when the change must stay coordinated across multiple records and you want one review + confirm step.
• Ask for a summary after every two or three staged changes so you can keep the delta understandable.
• Use Astro for the review narrative, risk framing, and stakeholder-ready summary before you move into the manual Promote flow.