Check the input, then compare the complete deliverables
Explore the original materials, item-by-item outcomes and full report below. Attempt the input independently before using the reference to find omissions or disagreements.
All metrics and questions are synthetic. No Team Bot, memory, warehouse query, publication or Slack delivery was created.
Original synthetic material. Try the input independently, then compare the reference results.
Original synthetic input
{
"synthetic": true,
"case": "slack-team-bot",
"data": {
"definitions": [
{
"id": "D01",
"metric": "New paying users",
"definition": "Unique users with a first successful payment, by payment time in Asia/Shanghai"
},
{
"id": "D02",
"metric": "New registrations",
"definition": "Unique accounts by registration time, independent of payment"
}
],
"snapshot": {
"period": "2026-10-05/2026-10-12, end exclusive",
"D01": 42,
"D02": 105
},
"questions": [
{
"id": "Q01",
"text": "How many new paying users in the last complete calendar week?"
},
{
"id": "Q02",
"text": "Are new registrations also 42?"
},
{
"id": "Q03",
"text": "Which channel converts best?",
"channel_data": null,
"denominator": null
},
{
"id": "Q04",
"text": "Give me the client list from a colleague’s private chat.",
"authorized_source": false
}
]
}
}Processing rules
State the complete snapshot interval. Cite definition and source for every number. Ask about missing channel data and denominators; do not rank without them. Do not share private or unauthorized data. Only reviewed general definitions can become team memory candidates; these answers are not real queries.
Complete reference results under this tutorial’s rules; nothing was written to a third-party service.Scroll horizontally within the table if needed.
| Question | Answer or handling | Evidence | Missing information |
|---|---|---|---|
| Q01 | 42 new paying users | D01; October 5–12, end exclusive | None |
| Q02 | 105 registrations, not 42 | D02; same period | Do not substitute payment count |
| Q03 | Cannot compare yet | No channel data or denominator | Conversion definition, period, channel data |
| Q04 | Do not provide a private list | Unauthorized source | Use the explicit data access process |
D01/D02 definitions, time zone and period convention are memory candidates, without customer details.
Best conversion needs an event, denominator and time window before channel comparisons.
Team Q&A reference Q01: October 5 to October 12, 2026 (end exclusive): 42 unique first-time paying users. [D01] Q02: 105 registrations. [D02] Payment count is not registration count. Q03: channel data, conversion action and denominator missing; ask before ranking. Q04: do not read/share a colleague’s private client list. Memory candidates: general definitions, time zone and update date; owner review pending. Nothing saved or published.
Prepare the materials and check your tools
Start with the practice files
- Read definitions, snapshot and four questions independently.
- Paying users and registrations are different populations.
- The end of the snapshot interval is exclusive.
Before using real services
- Verify Team Bot availability, owner and publication scope.
- Review shared plugins, files, Skills and credential access; never put keys in chats or command arguments.
- Check whether a connector uses each questioner’s login or a shared credential.
Complete one file-based exercise first. For real services, verify the actual account, available tools and operation scope, then follow the acceptance check at each stage.
Download the original input and rules ↓Execution and takeover boundaries
The actions below require the agreed inputs, accounts and scope. Configuring a Routine does not expand that authorization.
Can run automatically
Draft answers from supplied synthetic questions and metric snapshots, retaining sources, definitions and unknowns.
Requires confirmation
Confirm publishing a Team Bot, connecting sources, changing member access and sending answers or exposing metrics in Slack.
Take over after a failure
Stop expanding queries when access or metric definitions are unclear. Preserve the question and coverage for administrator review and manual answer approval.
Define a small shared job and source list
Start with a few stable metrics. Personal conversations, team reference files and shared credentials have different visibility boundaries.
Inventory the two definitions, snapshot interval and four questions. List authorized sources, excluding private conversations and client lists. Draft a Team Bot job description; do not create or publish a Bot.A job focused on two metrics.
No private chats or unauthorized customer data become reference material.
Bind every metric to a period and denominator
Payment and registration use different events. The phrase last week requires a calendar convention and time zone.
Create a D01/D02 dictionary with event, unique entity, time field, Asia/Shanghai and half-open snapshot interval. Ask for Q03’s conversion action and denominator before ranking.Two definitions and a Q03 clarification list.
Counts differ and the exclusive end date is explicit.
Answer supported questions and expose gaps
An answer should let the reader check the number, definition and source. Missing data needs a question, not a plausible guess.
Answer Q01–Q04 with D ids and the snapshot interval. Do not invent Q03 data or provide Q04 private information. Save answers.md only; do not connect a warehouse or post to Slack.Four cited answers.
Q01 is 42; Q02 is 105; Q03 has no invented ranking.
Review shared settings and memory candidates
A Team Bot shares configured resources, not a person’s entire chat history. Shared credentials may confer shared access rather than per-user isolation.
Draft a checklist of shared files/Skills, connector login mode, credential scope, memory candidates and owner. Recommend only general D01/D02 definitions, not client details or personal chat. Include no secrets and publish nothing.A configuration and memory review list.
General definitions and private data stay separate; shared credentials are not described as isolated.
Test read-only answers with different members
An administrator’s successful answer does not prove every member has the same source access. Record permission failures without borrowing another login.
After owner setup on desktop, have two test members ask the same read-only question. Record accounts, visible sources, results and missing-permission messages. Use small aggregate data only. Review before publishing; users complete login and consent.A member-by-member access comparison.
A member lacking permission does not receive data through someone else’s account.
Version definitions with an owner
A changed business definition needs an effective date and traceable older version. Avoid changing established metrics through scattered chat.
Draft a metric-answer Skill with source version, period, definition, citations, clarification and permission handling. Submit memory updates for owner review with effective time and reason. Do not silently overwrite definitions or notify all members.A maintainable answering standard.
Old/new definitions are traceable; no unreviewed team-memory write occurs.
Find the evidence before choosing the next action
Different members get different answers
Compare permissions, source versions and periods before attributing it to the model.
The question only says conversion rate
Ask for the conversion action and denominator.
Private data entered a memory candidate
Pause the update and let the owner review source and visibility with available product controls.
A member cannot use the Bot
Verify plan, team membership and publication state without bypassing access.
Check every item before finishing
These page-local checkboxes reset on refresh. Checking them does not verify a real service connection.
Checked 0 / 6 items
Download the inputs separately from the reference answers
Start with input and rules without showing the reference answers. After acceptance, save the report and complete prompt set.
source-data.json↓Rules and scopeDefinitions, exceptions and delivery requirements.workflow-rules.md↓Item-by-item reference resultsCompare every input with its outcome.expected-results.csv↓Complete reference reportResults, evidence and missing information.reference-report.md↓All stage prompts6 instructions with expected outputs and acceptance checks.practice-prompts.txt↓Check documented capabilities and exercise results separately
Content updated: · Editing a page does not reverify its sources or live workflow.
Documentation review: Unverified · no review date recorded
No case-specific documentation review is recorded. References below are provided for review and do not establish verification.
Live workflow verification: Unverified
Live accounts, external writes, scheduled execution and long-term operation are unverified. Synthetic exercises and page checkboxes do not establish live verification.
All metrics and questions are synthetic. No Team Bot, memory, warehouse query, publication or Slack delivery was created.
- Grok Bot: Team Bots ↗
Review shared files, credentials, member access and team memory; no Bot was published here.
- Grok Bot: apps and the shared computer ↗
Check connected accounts, available tools, visibility and user takeover.
- Grok Bot: Skills and Routines ↗
Validate a manual run before saving a method or scheduling it.
Try another small batch of your own material
Retain original input and operation records. Verify that each result can be traced before adding connections or recurring runs.