Design Bounded Collection and Run Controls

What you will complete

Complete a reviewed Collection Configuration and Run Manifest with exact scope, caps, timeout, charge ceiling, and stop behavior.

Decision rule: The browser or prompt never gets to weaken a server-side cap, provider allowlist, or prohibited action.

Show the exact run before it starts

The operator should see the query or URL set, provider or retrieval method, source scope, result cap, timeout, disabled add-ons, destination, and maximum charge. Validate duplicated or malformed targets before a billable call. A schedule does not remove the need for an approved configuration.

Preserve control at the execution layer

Apply source allowlists, item caps, timeout, and charge ceiling where the collection actually runs. Do not rely on form controls or prompt instructions. Never retry a paid run automatically. Authentication, billing, policy, quota, timeout, and unexpected schema errors should stop with a visible status.

Keep secrets and private data out

Use an ignored local environment file or an approved secret manager. Never place tokens in browser code, lesson files, reports, fixtures, logs, or exports. This course uses public sources. Private URLs, cookies, custom authorization headers, personal-data enrichment, CRM writes, messaging, and publishing require a separate approved design.

Worked example: Northstar Workflow Studio

Fictional teaching fixture: Northstar's fictional configuration contains one reviewed query, one page per query, a ten-record normalization ceiling, a ten-minute timeout, no paid add-ons, and a small explicit charge limit. The run manifest records a local run identifier, configuration hash, collection start, source-policy version, rubric version, and status before provider work begins.

Northstar is not a Flowgrammer client, and its sources, weights, records, scores, timing, and results are not live evidence. The example exists so you can inspect how one choice changes the same capstone from lesson to lesson.

Workbook exercise

Open section 3 of the Morning Intelligence Agent Launch Pack. Complete the following work for your own approved topic:

1. Write the exact query or URL set and approved provider or retrieval method.
2. Set item, page, time, and cost caps.
3. List every disabled add-on and prohibited data source.
4. Define stop behavior for authentication, billing, policy, quota, timeout, empty output, and schema drift.

Write assumptions as assumptions. Keep any untested field marked untested. Ask another person to challenge one rule before you continue.

Common failure modes

• An unbounded query hidden behind a simple Run button.
• Automatic retry after a paid timeout.
• A token copied into a report for convenience.
• Treating a scheduled run as pre-approved scope expansion.

Repair the smallest responsible layer: scope, source policy, collection control, evidence model, ranking rule, report, or operating gate. Do not hide a failed condition inside a more confident summary.

Capstone connection

Complete a reviewed Collection Configuration and Run Manifest with exact scope, caps, timeout, charge ceiling, and stop behavior. This section must be specific enough for another operator to inspect, test, and reject. A complete worksheet is not automatically a passing worksheet; the evidence and critical gates still control the decision.

Check your application