User reference / Read the reports
Report orientation & About
Read the run identity, recorded versions, council, analysis history and shared controls.
| Field | Source and interpretation |
|---|---|
| IdReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | task.task_id, identifying one execution. |
| PromptReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Human label plus content-derived prompt_key; the label links to prompt history. “Not recorded” is possible in old traces. |
| ConfigurationReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Saved label and configuration_key. Different labels produce different configuration keys even with the same model settings. |
| Loaded fromReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Base-file basenames in loading order, deepest inherited file first. These are provenance, not clickable configuration contents; no credentials are shown. |
| Made withReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Executor version, package release and trace format recorded for the task. Missing old version fields are shown as unrecorded. |
| This pageReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Current renderer version plus the selected stored analysis's generator/schema/package when available. |
| StatusReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Stored task status; consult execution errors for why a run stopped. |
| StartedReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Task creation timestamp shown to minute precision. |
| DurationReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Task update timestamp minus creation timestamp. A malformed/incompatible timestamp pair gives N/A; negative differences are clamped to zero. Old timezone records have limitations. |
| Field | Source and interpretation |
|---|---|
| ProposersReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Number of stored round-0 proposals. It describes recorded participation, not necessarily the requested count. |
| Refinement roundsReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Highest recorded round number; round 0 is initial. |
| VotersReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Number of valid saved votes. |
| Selected planReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Winning proposal's agent, votes received / valid votes and selected step count. |
| Plan costReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Estimated sum of stored proposal and voting calls. Without complete pricing, the card can show tokens instead. The small text reports calls, input and output tokens. |
| Analysis cardReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Optional selected assessment's analyst name/provider/model/settings, calls and estimated cost or token/availability information. It is separate from collaboration cost. |
| Models of the councilReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Distinct model descriptions from proposal records and their effective thinking/sampling/output settings. It is not simply the currently configured JSON list. |
For how usage is collected and how prices are applied, see Timeline & costs. A stored token zero can mean a provider did not report the detail. It does not prove that no computation occurred.
| Column | Meaning |
|---|---|
| #Reviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Display position in this run's saved analysis history. It is not the analysis UUID or a quality rank. |
| DateReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Analysis creation time. This can be much later than the execution. |
| AnalystReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Saved name, provider/model and effective settings used for that assessment. |
| SchemaReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Structure version of the stored analysis content. This differs from the analyst component version. |
| Analysis generatorReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Component version and package release recorded on the assessment. |
| CallsReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Length/usage of that assessment's recorded call list. Old records can lack usage. |
| CostReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Estimate for that analysis alone using the current price list. |
| Verdict on the voteReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Saved agreement/disagreement where the schema supports it; absent otherwise. |
| ReportReviewed E 0.1.5 · A 0.1.4 · R 0.1.16 | Open the page displaying that analysis, or “this page” for the current one. |
report.html uses the newest stored analysis. Earlier pages reuse the same execution trace with a different assessment. They are not new council runs. About ends with the full task statement rendered as Markdown.
White/general report surfaces contain recorded output or deterministic calculations. Yellow LLM analysis surfaces contain the analyst's interpretation. Blue declared by the agents annotations describe legacy notes. The explicit computed by the report tool tag identifies title-based calculations.
For elements with explanatory popovers, hover or focus to read, click to pin, and click the close mark, outside the popover or Escape to dismiss. A pinned popover can be dragged by its header. Following a fragment link closes it. These details make a number inspectable; their presence does not turn inferred influence into causal proof.
Source files used for this reference
Reviewed at 7361ea0b71a5.
