Context Boundary Mapper
A practical worksheet for deciding what an AI system can see, remember, retrieve, and produce before implementation begins.
Fill this onlineTemplate preview
The exact sections inside the download.
Frame the workflow
Map the operating context before choosing tools. Name the workflow or decision supported, the specific operating moment, and the human owner accountable for quality, exceptions, and escalation.
Map authoritative context
List the systems of record and the model or context service that may retrieve them. Distinguish authoritative sources from convenience copies and state what may be searched, summarized, or cached.
Set interface and governance
Show where people see, approve, edit, or reject AI output, and where permissions, logs, retention rules, and review gates are enforced.
Classify permitted context
Separate approved operating sources from restricted information and no-retention inputs. Record the purpose, access boundary, and expiry rule for each context class.
Operate memory deliberately
If the system remembers or retrieves context, name its owner, authorized reuse rule, revision cadence, and evidence trail.
Trace every decision
Connect the input source, retrieval rule, output consumer, evidence log, and human review trigger so the team can explain what happened and why.
01 / 03
Frame the decision
Name the real operating need before designing a solution.
Workflow or decision supported.
Name the specific operating moment this AI system supports.
Name the role accountable for quality, exceptions, and escalation.
List the authoritative sources. Do not include convenience copies.
How to use it
Start with one real decision.
Complete the canvas with the workflow owner, then use the blank areas to expose missing context and controls.