Create a blank Copilot Studio agent
Step 1 of 26
Action
Create a blank Copilot Studio agentExpected resultA blank Copilot Studio agent is visible in the captured Draft workspace.
Manual mode · literal browser construction
Build Maintenance Scheduling Agent manually.
No PAC CLI, YAML import, or plugin architect. Follow each action and its evidence boundary, compare only the reviewed references, and stop at Draft.
Synthetic disclosure: this is qualitative workflow evidence using packaged synthetic inputs. It is not a customer KPI or a live-system result.
Progress will remain available on this page, but it will not persist after you leave.
Found something inaccurate? Use Report an issue on that step. It opens a prefilled GitHub issue for review and does not submit automatically.
Build and verify
Name Maintenance Scheduling Manual
Step 2 of 26
Action
Name Maintenance Scheduling ManualMaintenance Scheduling Manual
Expected resultThe page header shows the recorded manual build name: Maintenance Scheduling Manual.
Enter the reviewed global instructions
Step 3 of 26
Action
Enter the reviewed global instructions# Maintenance Scheduling Agent — Manual Global Instructions Use only the uploaded synthetic records, review rules, and operation skills. Treat every exact figure, identifier, name, date, score, duration, and cost as synthetic pilot evidence. ## Boundaries - State that the source is a fixed synthetic snapshot and do not imply live access. - Do not browse or invent records, actions, confirmations, or outcomes. - Do not narrate internal retrieval, tool selection, restrictions, or implementation mechanics; present only the user-facing result. - The agent never controls equipment or creates, assigns, schedules, or dispatches maintenance work. - Recommend the approved human review and production connection required for any action. ## Response contract Lead with the relevant record and evidence, distinguish facts from recommendations, name the authorization gate, and state that no external side effect occurred. The fixed scenario is Line 3 injection molding: Machine #7 (78% screw wear, 120 operating hours to failure) before a 50,000-unit order. Route machine condition to `predictive-alerts`, "show me the schedule" to `work-order-plan` (Saturday 6 AM - 10 AM, 104% capacity margin), the cost analysis to `downtime-analysis` ($3,200 vs $18,500, ROI 478%), "schedule everything" to `work-order-draft` (draft WO-2024-3847 for approval), the 30-day calendar to `maintenance-calendar`, and long-term recommendations to `optimization-plan`. Machine #7 is the default machine. <!-- locked-preview-anchors:start --> ## Skill routing map Route from the user's natural-language intent to the correct skill below. Do not narrate internal retrieval, tool selection, restrictions, or implementation mechanics; present only the user-facing result. - `MS-01` uses skill `schedule-overview`. - `MS-02` uses skill `predictive-alerts`. - `MS-03` uses skill `work-order-plan`. - `MS-04` uses skill `downtime-analysis`. - `MS-05` uses skill `work-order-draft`. - `MS-06` uses skill `maintenance-calendar`. - `MS-07` uses skill `optimization-plan`. These skill names above are the ONLY valid skill identifiers. Never invent, guess, or reference any other skill name. If the correct skill or its knowledge cannot be loaded after one retry in the same turn, say so honestly and stop. Do not answer using values you already know from these instructions or from general knowledge -- a response with no real citation is not acceptable output. <!-- locked-preview-anchors:end -->
Expected resultThe reviewed manual/GLOBAL-INSTRUCTIONS.md policy is visible or saved without unrecorded edits.
Save the instruction policy
Step 4 of 26
Action
Save the instruction policyExpected resultThe reviewed manual/GLOBAL-INSTRUCTIONS.md policy is visible or saved without unrecorded edits.
Remove default web search
Step 5 of 26
Action
Remove default web searchExpected resultThe captured inventory no longer lists the default web-search capability.
Add knowledge: aibast_maintenance-scheduling-review-rules.md
Step 6 of 26
Action
Add knowledge: aibast_maintenance-scheduling-review-rules.mdExpected resultThe exact knowledge file linked for this step is attached under its complete filename. Wait for processing to finish before continuing. This step does not require the final inventory or prove the other uploads.
Add knowledge: aibast_maintenance-scheduling-synthetic-records.md
Step 7 of 26
Action
Add knowledge: aibast_maintenance-scheduling-synthetic-records.mdExpected resultThe exact knowledge file linked for this step is attached under its complete filename. Wait for processing to finish before continuing. This step does not require the final inventory or prove the other uploads.
Add skill: aibast_downtime-analysis_04
Step 8 of 26
Action
Add skill: aibast_downtime-analysis_04Expected resultThe linked SKILL.md appears with its frontmatter name and description. Open it and compare the complete instructions with the source file. This step does not require the final inventory or prove the other uploads.
Add skill: aibast_maintenance-calendar_06
Step 9 of 26
Action
Add skill: aibast_maintenance-calendar_06Expected resultThe linked SKILL.md appears with its frontmatter name and description. Open it and compare the complete instructions with the source file. This step does not require the final inventory or prove the other uploads.
Add skill: aibast_optimization-plan_07
Step 10 of 26
Action
Add skill: aibast_optimization-plan_07Expected resultThe linked SKILL.md appears with its frontmatter name and description. Open it and compare the complete instructions with the source file. This step does not require the final inventory or prove the other uploads.
Add skill: aibast_predictive-alerts_02
Step 11 of 26
Action
Add skill: aibast_predictive-alerts_02Expected resultThe linked SKILL.md appears with its frontmatter name and description. Open it and compare the complete instructions with the source file. This step does not require the final inventory or prove the other uploads.
Add skill: aibast_schedule-overview_01
Step 12 of 26
Action
Add skill: aibast_schedule-overview_01Expected resultThe linked SKILL.md appears with its frontmatter name and description. Open it and compare the complete instructions with the source file. This step does not require the final inventory or prove the other uploads.
Add skill: aibast_work-order-draft_05
Step 13 of 26
Action
Add skill: aibast_work-order-draft_05Expected resultThe linked SKILL.md appears with its frontmatter name and description. Open it and compare the complete instructions with the source file. This step does not require the final inventory or prove the other uploads.
Add skill: aibast_work-order-plan_03
Step 14 of 26
Action
Add skill: aibast_work-order-plan_03Expected resultThe linked SKILL.md appears with its frontmatter name and description. Open it and compare the complete instructions with the source file. This step does not require the final inventory or prove the other uploads.
Review model, skills, knowledge, and safety boundaries
Step 15 of 26
Action
Review model, skills, knowledge, and safety boundariesExpected resultThe screenshot visibly confirms Claude Opus 5, 7 rendered skills, and no tools. Knowledge is verified separately in the two knowledge-upload checkpoints.
Review the skill list (7 skills)
Step 16 of 26
Action
Review the skill list (7 skills)Expected resultThe screenshot visibly confirms Claude Opus 5, 7 rendered skills, and no tools. Knowledge is verified separately in the two knowledge-upload checkpoints.
Review the rest of the skill list
Step 17 of 26
Action
Review the rest of the skill listExpected resultThe screenshot visibly confirms Claude Opus 5, 7 rendered skills, and no tools. Knowledge is verified separately in the two knowledge-upload checkpoints.
Open a fresh Preview conversation
Step 18 of 26
Action
Open a fresh Preview conversationExpected resultA fresh Preview surface or its recorded qualitative result is visible; only the evidence file defines a pass.
Pass MS-01: Maintenance Manager
Step 19 of 26
Action
Pass MS-01: Maintenance ManagerGive me the equipment and technician-capacity picture I should review before the weekly maintenance meeting.
Expected resultThe captured Preview evidence records MS-01 with the recorded identifiers IM-07, Technician Availability, Marcus Chen; do not infer results beyond it.
Before this case
Open Preview in a fresh conversation before running MS-01. A previous response must not influence the evidence.
Pass MS-02: Production Supervisor
Step 20 of 26
Action
Pass MS-02: Production SupervisorMachine #7 on Line 3 is showing wear indicators and we have a big order next week. What does the condition data say?
Expected resultThe captured Preview evidence records MS-02 with the recorded identifiers Screw wear at 78%, 120 operating hours, 50,000 units; do not infer results beyond it.
Before this case
Open Preview in a fresh conversation before running MS-02. A previous response must not influence the evidence.
Pass MS-03: Maintenance Manager
Step 21 of 26
Action
Pass MS-03: Maintenance ManagerShow me the lowest-impact maintenance window for machine #7 with coverage, crew, parts and backup, but do not create or dispatch work orders.
Expected resultThe captured Preview evidence records MS-03 with the recorded identifiers Saturday 6 AM - 10 AM, 104%, No work order is created or dispatched; do not infer results beyond it.
Before this case
Open Preview in a fresh conversation before running MS-03. A previous response must not influence the evidence.
Pass MS-04: Operations Leader
Step 22 of 26
Action
Pass MS-04: Operations LeaderWhat does preventive maintenance on machine #7 cost compared with a breakdown, and what is the ROI?
Expected resultThe captured Preview evidence records MS-04 with the recorded identifiers $3,200, $18,500, 478%; do not infer results beyond it.
Before this case
Open Preview in a fresh conversation before running MS-04. A previous response must not influence the evidence.
Pass MS-05: Maintenance Manager
Step 23 of 26
Action
Pass MS-05: Maintenance ManagerLine up the crew and parts for the machine #7 overhaul so I can approve it.
Expected resultThe captured Preview evidence records MS-05 with the recorded identifiers WO-2024-3847, Bin 12-A, No work order is created or dispatched; do not infer results beyond it.
Before this case
Open Preview in a fresh conversation before running MS-05. A previous response must not influence the evidence.
Pass MS-06: Operations Leader
Step 24 of 26
Action
Pass MS-06: Operations LeaderWhat does the maintenance calendar look like for the next 30 days?
Expected resultThe captured Preview evidence records MS-06 with the recorded identifiers Machine #12, 87% planned, Calibration; do not infer results beyond it.
Before this case
Open Preview in a fresh conversation before running MS-06. A previous response must not influence the evidence.
Pass MS-07: Operations Leader
Step 25 of 26
Action
Pass MS-07: Operations LeaderWhat long-term changes would cut our maintenance cost and unplanned downtime?
Expected resultThe captured Preview evidence records MS-07 with the recorded identifiers Quick Wins, $124K, Run pilot on Line 3; do not infer results beyond it.
Before this case
Open Preview in a fresh conversation before running MS-07. A previous response must not influence the evidence.
Confirm Draft and stop before publish
Step 26 of 26
Action
Confirm Draft and stop before publishExpected resultThe agent remains Draft and no Publish action is taken.
Troubleshooting
A screenshot or browserfilm frame is missing
Do not invent, recreate, or substitute an image. Keep the step open until its actual evidence is captured and reviewed. Remove --allow-pending only after all workshop evidence gates are met.
A knowledge file is still processing
Wait for ingestion to finish before Preview. A partial answer is not evidence.
A skill upload fails
Use the linked raw SKILL.md. Fix the reviewed source deliberately; do not silently skip the action.
The model differs from Easy mode
Record the substitution and stop the parity claim until Easy and Manual use the same reviewed model.
The Preview answer misses an identifier
Mark the recorded case failed, inspect instructions and inventory, then replay the exact prompt in a fresh conversation.
Should I publish?
No. Keep the same owned manual agent in Draft. Do not choose Publish as part of this tutorial.