Reliable recurring work
Steps, handoffs, failures, and tool boundaries
Design a sequence whose intermediate results, stop conditions, and ownership can be checked.
Lesson 2 of 6 in the recommended order · About 25 min (estimate)
On this page
Practical AI glossary — terms and common confusions
- Model
- A learned component used to produce a result.
A drafting model is one part of a letter app; it is not the whole app.
- Application
- The software experience around components and services.
A letter app adds accounts, storage, and a Send button.
- Prompt
- Instructions and input supplied for a task.
“Summarize this notice in two bullets” specifies a task and shape, not a truth guarantee.
- Token
- A unit a model processes; it need not be a whole word.
A tokenizer can split a name into pieces; count using the actual system.
- Context
- Information available for the current request.
An earlier attachment may be absent even when its filename is visible.
- Training
- Adjusting a model using examples.
A training log differs from a conversation correction.
- Inference
- Using a trained model to produce a result.
Drafting another sentence does not by itself prove retraining.
- Retrieval
- Selecting material to supply to a task.
A retrieved notice still needs a relevance and date check.
- Embedding
- A numerical representation used to compare or organize items.
A similar search match is a candidate, not proof of the same meaning.
- Tool
- A function an application can call.
A calendar lookup can be read-only; finding an event is different from creating one.
- Agent
- Software that can choose and carry out steps toward a goal.
An agent still needs bounded permissions and checked action receipts.
- Hallucination
- Plausible generated content that is unsupported or false.
An invented opening time needs evidence or removal, even when phrased confidently.
- Evaluation
- Checking results against stated criteria.
Twenty checked examples establish a bounded result, not universal reliability.
- Multimodal
- Working with more than one kind of input or output.
A photo plus text does not prove every small label in the photo was read correctly.
Outcome
Use workflow receipts to identify the safe next step after a partial result or failure.
Why it matters
Design a sequence whose intermediate results, stop conditions, and ownership can be checked.
Concept
A recurring workflow needs more than a list of desired outputs. For each step, define input, owner or tool, expected result, verification, and failure behavior. A later step should not silently proceed when a required earlier result is absent.
Handoffs should identify the last confirmed state and what remains. Avoid rerunning a whole workflow just because the final step failed: earlier actions may already have taken effect. For uncertain external actions, check their status before retrying. For failed reads, recover the missing input and keep incomplete results clearly labeled.
Worked example
A fictional workflow reads a roster, prepares invitations, obtains approval, then sends. If approval is pending, the correct next state is “draft ready; approval pending,” not “sent.” If sending partially succeeds, identify which recipients have confirmed receipts before planning any retry.
Planned workflow: each handoff needs evidence
If approval is pending: draft ready; approval pending.
Stop before sending. A prepared draft does not cross the human approval boundary.
1. Read roster
A missing or failed read leaves required input absent. Recover that input rather than treating an incomplete roster as complete.
2. Prepare invitations
Prepared invitations are drafts. Preparation alone is not approval and does not establish that anything was sent.
3. Obtain approval
Human approval is the boundary before sending. Use the authorized scope; do not silently proceed while approval is pending.
4. Send and inspect receipts
If results are partial, retain each recipient’s confirmed status. Check uncertain outcomes before planning a retry; do not rerun every earlier action.
Before you check
Write a four-step workflow table with Input, Action, Verification, and On failure. Include one human approval boundary and one recovery check.
Practice and fresh transfer
The packets below are fictional and contain the facts needed for these cases. External references are optional background. Judge each response independently: it may be supported, contradicted, or unresolved. Select the passages needed to justify your judgment and write why the distinction matters before revealing feedback.
Assess the response
Use Accept when all material claims are supported. Use Revise when a supplied fact or requirement is contradicted. Use Evidence is insufficient when a key fact cannot be established either way. If a response contains both an unknown and a direct contradiction, choose Revise and explain both problems. Conflicting claims with no established authority remain insufficient; a claim does not become a governing fact merely because a source asserts it.
Some responses are fully supported. Others need correction or more evidence. Judge each on its sources; do not edit a correct answer just to change it.
Your written notes stay in this page and disappear when you leave. Only a self-reviewed completion can be saved to your learning account. These practice checks do not establish independent proficiency.
Practice
Case 1
Consider a fictional workflow handoff.
Source packet
- Source 1
- Workflow: Read schedule, draft invitation, obtain coordinator approval, then send.
- Source 2
- Log: Schedule read; draft ready; approval not yet received.
Response to assess
The draft is ready for coordinator approval; sending is not yet authorized in this workflow.
Practice
Case 2
Consider a fictional retry plan after partial delivery.
Source packet
- Source 1
- Receipts: Messages to A and B were delivered. Message to C failed before sending.
- Source 2
- Retry rule: Retry only confirmed failed recipients; do not resend confirmed deliveries.
Response to assess
Resend all three messages from the start.
Fresh transfer
Case 3
Consider a fictional read failure in a comparison workflow.
Source packet
- Source 1
- Read log: File A loaded; file B failed. No output comparison has been produced.
- Source 2
- Workflow rule: A complete comparison requires both files.
Response to assess
Recover file B before producing the complete comparison; an A-only summary would need an explicit scope label.
Fresh transfer
Case 4
Consider a fictional downstream action after a failed validation.
Source packet
- Source 1
- Validation receipt: The total does not match the source rows.
- Source 2
- Workflow rule: Do not publish the report until totals match and the reviewer approves.
Response to assess
Publish now and repair the total afterward.
Save your self-review
Completion records that you reviewed the cases. Your explanation and transfer performance need a facilitator to establish independent learning.
Sign in with your learning-center account to save completion.
Review the explanation for every case before saving.
Summary and next step
Apply the checklist to a new task. Preserve supported content, explain any change with evidence, and name what remains unresolved. Saving records self-review, not independently demonstrated proficiency. A facilitator must assess the explanation and fresh transfer for human learning evidence.