SpyderBot Documentation

Reports, agent usage and action plans

Reports and connected agents use the same metric definitions and evidence semantics as the product. A different surface does not create a different meaning for “visibility”. Access permissions still apply to prompts, answer text, citations, source URLs and other underlying evidence.

A report is a snapshot with its own period, scope, method version and generation time. A generation timestamp is not a claim that all underlying sources were observed at that instant.

On this page

Report Scope and Freshness · Agent Quotes, Credits and Usage · Action Plan Progress

Report Scope and Freshness

Unit: Scope and timestamps

The measurement identity and time context needed to interpret a report.

How it is assessed

The report pins its question and entity scope, systems, channels, observation window, aggregation profile, method version and source timestamps. Observed, estimated and diagnostic values retain their own provenance. A cached snapshot keeps its original observation context when reopened.

How to interpret it

Reopening a report does not make its evidence current. Exported values must preserve their original scope and availability states.

What to do next

Use the report’s metric explanation link and verify its version before comparison.

Measurement scope and aggregation ↗ · Evidence and availability ↗

Agent Quotes, Credits and Usage

Unit: Quoted and recorded billing units

The displayed resource or credit estimate for an agent-requested operation and its recorded usage.

How it is calculated

Use the current operation quote and account-specific usage record as the authority. Distinguish a quote, an authorised operation, work in progress and recorded final usage. Do not infer customer billing from model tokens, HTTP retries or UI execution counts. Plan capabilities and prices come from current product configuration.

How to interpret it

Credits measure usage or billing, not measurement quality. Additional credits do not automatically increase a prompt’s weight in a headline.

What to do next

Review the operation scope and quote before the relevant billable action; reconcile usage with the recorded operation.

Measurement scope and aggregation ↗ · Evidence and availability ↗

Action Plan Progress

Unit: Task progress and verification state

Work completed on a recommended action plan and evidence that the intended change was verified.

How it is assessed

Task progress counts the actual statuses of the tasks in the declared plan. Implementation verification, real bot activity and measured outcome change are separate evidence milestones. Available supporting findings explain the recommendation within the viewer’s access permissions.

How to interpret it

A completed plan is not proof of higher AI visibility, traffic or revenue.

What to do next

Verify the implementation and evaluate comparable follow-up measurements.

Related concepts: Remediation Verification Status ↗ · Metric Change ↗

Measurement scope and aggregation ↗ · Evidence and availability ↗