Show reported provider spend in CLI dashboards

[Chipagosfinest]

OpenRouter Activity already supplies reported spend in its usage snapshot, but the CLI dashboard emitted cost: null because its producer only projected the separate local-cost result. Dashboard-v1 now uses a provider's compatible 30-day USD aggregate when no local cost payload exists. Existing local results retain precedence, including an unavailable local amount.

Provider history keeps its own calendar meaning. Completed UTC daily buckets never become local Today merely because their date string matches: todayUSD stays null, known zero stays zero, and incomplete-request counts remain attached to the reported window. Other currencies or history lengths remain unavailable in the fixed USD/30-day fields.

Validation

• make check and independent P0–P2 review passed.
• All 1,247 selections across 114 full-suite groups passed with no failures, retries, or timeouts.
• 27 focused tests passed under TZ=America/Los_Angeles, including the UTC/local rollover, known zero, missing totals, incompatible currencies/windows, incomplete history, and local-cost precedence.
• Built, signed CLI before/after proof used identical no-network synthetic provider fixtures in task-owned copies of the helper and its resource bundle. Real-home reads/writes and all network access were denied for the one-shot command. Both exited successfully: baseline cost: null; candidate {"todayUSD":null,"last30DaysUSD":12.5}.
• The authenticated loopback HTTP route returned the same result, and the real Chrome dashboard rendered Last 30 days $12.50 with Today omitted. The fixture makes no provider API requests and uses no real credentials.
• CI run 35484141518 passed all nine checks on a8ee4c977986afd3c2bf086a78e9f4672035585b, including macOS, Linux, and GitGuardian.

Live dashboard before and after

The screenshots contain only synthetic data.

Before

After

Thanks @Chipagosfinest for identifying the missing dashboard projection.

[clawsweeper]

🦞👀
ClawSweeper picked this up.

Pull request received. I will update this pull request when review starts.

ClawSweeper review complete

ClawSweeper finished reviewing this revision. The review result is being finalized.

View the workflow run.

[Chipagosfinest]

Live 0.61.0 check that motivated this: codexbar usage --provider openrouter already prints Last 30 days (UTC): $… (reported) plus Activity token/request/model counts. codexbar dashboard left cost: null for the same provider because only the local cost-command providers were attached.

@clawsweeper re-review

[clawsweeper]

🦞🧹
ClawSweeper re-review requested.

I asked ClawSweeper to review this item again.
Action: item re-review queued (workflow sweep.yml, event exact_review_queue).
Result: when the review finishes, ClawSweeper will create the durable review comment if needed or update the existing comment in place.

[clawsweeper]

Codex review: needs maintainer review before merge. Reviewed September 20, 2026, 1:10 AM ET / 05:10 UTC (Revision 4).

ClawSweeper review

What this changes

Adds provider-reported 30-day USD spend to CLI dashboard snapshots when no local cost result exists, with regression tests and documentation.

Merge readiness

✅ Ready for maintainer review

The PR remains useful: current main and v0.62.0 still omit provider-reported dashboard spend. The revised scope and new before-and-after evidence address the prior review; no blocking defects remain.

Priority: P2
Reviewed head: a8ee4c977986afd3c2bf086a78e9f4672035585b

Review scores

Measure	Result	What it means
Overall readiness	🐚 platinum hermit (4/6)	A focused, compatible implementation with direct before-and-after proof and useful regression coverage.
Proof confidence	🦞 diamond lobster (5/6) ✨ media proof bonus	Sufficient (screenshot): The captured built-CLI before/after output and inspected Chrome screenshots exercise the dashboard projection with isolated synthetic provider inputs: the aggregate becomes $12.50 and Today remains absent. This proves the changed display behavior without requiring live provider credentials; upstream fetching and browser security behavior are unchanged.
Patch quality	🐚 platinum hermit (4/6)	No actionable review findings were identified.

Verification

Check	Result	Evidence
Real behavior	Verified	Sufficient (screenshot): The captured built-CLI before/after output and inspected Chrome screenshots exercise the dashboard projection with isolated synthetic provider inputs: the aggregate becomes $12.50 and Today remains absent. This proves the changed display behavior without requiring live provider credentials; upstream fetching and browser security behavior are unchanged.
Evidence reviewed	7 items	Introduced change and compatibility: The pinned merge-base-to-head delta changes four files. The builder preserves any local cost payload, including unavailable local amounts; otherwise it accepts only USD/30-day history, preserves known zero and incomplete counts, and leaves Today unknown. Still needed on main and in the release: Inspected the builder at fetched main d987dd5 and release v0.62.0 (4b3ed1a). Both still call makeCost exclusively and return no cost when the separate local result is absent; neither contains this fallback. Provider calendar and consumer contract: OpenRouter’s bundled provider emits USD history for 30 completed UTC days. The existing web renderer independently displays available cost fields, so an unknown Today value is omitted while the aggregate appears. No provider fetch, credential handling, or HTTP authorization changes are introduced.
Findings	None	None.
Security	None	None.

How this fits together

CodexBar combines provider usage and local cost results into dashboard snapshots consumed by the CLI and authenticated web dashboard. This change lets the existing cost display use compatible spend already present in a provider’s usage snapshot.

flowchart LR
 A[Provider usage history] --> C[Dashboard snapshot builder]
 B[Local cost results] --> C
 C --> D{Local result exists?}
 D -->|Yes| E[Preserve local cost]
 D -->|No| F[Check USD and 30-day history]
 E --> G[CLI and web dashboard]
 F -->|Compatible aggregate with Today unknown| G

Loading

Before merge

None.

Agent review details

Security

None.

Review metrics

Metric	Value	Why it matters
Production and test delta	Production +15/-1; tests +122/-0	The small projection helper is justified by missing dashboard coverage and has focused boundary tests.

Technical review

Best possible solution:

Reuse compatible provider aggregates through the existing dashboard cost fields while preserving local precedence and unknown Today values.

Do we have a high-confidence way to reproduce the issue?

Not applicable as a new dashboard data-source capability; source inspection confirms the existing local-only limitation, and supplied before-and-after evidence demonstrates the added behavior.

Is this the best way to solve the issue?

Yes. The fallback reuses existing data and output fields without another fetch, competing cost pipeline, or invented local-day total.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning medium; reviewed against d987dd59ad34.

Labels

Label changes:

• add proof: sufficient: Contributor real behavior proof is sufficient. The captured built-CLI before/after output and inspected Chrome screenshots exercise the dashboard projection with isolated synthetic provider inputs: the aggregate becomes $12.50 and Today remains absent. This proves the changed display behavior without requiring live provider credentials; upstream fetching and browser security behavior are unchanged.
• add proof: 📸 screenshot: Contributor real behavior proof includes screenshot evidence. The captured built-CLI before/after output and inspected Chrome screenshots exercise the dashboard projection with isolated synthetic provider inputs: the aggregate becomes $12.50 and Today remains absent. This proves the changed display behavior without requiring live provider credentials; upstream fetching and browser security behavior are unchanged.
• add rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🦞 diamond lobster and patch quality is 🐚 platinum hermit.
• add status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (screenshot): The captured built-CLI before/after output and inspected Chrome screenshots exercise the dashboard projection with isolated synthetic provider inputs: the aggregate becomes $12.50 and Today remains absent. This proves the changed display behavior without requiring live provider credentials; upstream fetching and browser security behavior are unchanged.
• remove status: 📣 needs proof: Current PR status label is status: 👀 ready for maintainer look.
• remove rating: 🦪 silver shellfish: Current PR rating is rating: 🐚 platinum hermit, so this older rating label is no longer current.

Label justifications:

• P2: Displaying already-available provider spend is a bounded dashboard improvement without an urgent runtime failure.
• rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🦞 diamond lobster and patch quality is 🐚 platinum hermit.
• status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Sufficient (screenshot): The captured built-CLI before/after output and inspected Chrome screenshots exercise the dashboard projection with isolated synthetic provider inputs: the aggregate becomes $12.50 and Today remains absent. This proves the changed display behavior without requiring live provider credentials; upstream fetching and browser security behavior are unchanged.
• proof: sufficient: Contributor real behavior proof is sufficient. The captured built-CLI before/after output and inspected Chrome screenshots exercise the dashboard projection with isolated synthetic provider inputs: the aggregate becomes $12.50 and Today remains absent. This proves the changed display behavior without requiring live provider credentials; upstream fetching and browser security behavior are unchanged.
• proof: 📸 screenshot: Contributor real behavior proof includes screenshot evidence. The captured built-CLI before/after output and inspected Chrome screenshots exercise the dashboard projection with isolated synthetic provider inputs: the aggregate becomes $12.50 and Today remains absent. This proves the changed display behavior without requiring live provider credentials; upstream fetching and browser security behavior are unchanged.

Evidence

What I checked:

• Introduced change and compatibility: The pinned merge-base-to-head delta changes four files. The builder preserves any local cost payload, including unavailable local amounts; otherwise it accepts only USD/30-day history, preserves known zero and incomplete counts, and leaves Today unknown. (Sources/CodexBarCLI/DashboardSnapshotBuilder.swift:145, a8ee4c977986)
• Still needed on main and in the release: Inspected the builder at fetched main d987dd5 and release v0.62.0 (4b3ed1a). Both still call makeCost exclusively and return no cost when the separate local result is absent; neither contains this fallback. (Sources/CodexBarCLI/DashboardSnapshotBuilder.swift:145, d987dd59ad34)
• Provider calendar and consumer contract: OpenRouter’s bundled provider emits USD history for 30 completed UTC days. The existing web renderer independently displays available cost fields, so an unknown Today value is omitted while the aggregate appears. No provider fetch, credential handling, or HTTP authorization changes are introduced. (Sources/CodexBarCore/Resources/Plugins/openrouter.js:263, a8ee4c977986)
• After-fix production-path evidence: The captured PR body reports built CLI baseline/candidate runs using identical isolated synthetic provider fixtures: cost changed from null to an aggregate of 12.5 with Today null. It also reports the authenticated loopback response and real Chrome rendering. Locally inspected both downloaded attachments: https://github.com/user-attachments/assets/98f1273e-72f0-4f7e-9df6-ae2df5f5d842 shows no cost metric; https://github.com/user-attachments/assets/b5322f5c-7cdc-44a4-bc4f-e1e96a4bca46 shows Last 30 days $12.50 with Today omitted. This directly covers the changed projection and visible output. (a8ee4c977986)
• Regression coverage and review continuity: Read all 122 lines of the added tests, covering zero, missing totals, incompatible currencies/windows, local precedence, JSON null Today, and incomplete history. Earlier UTC concerns are resolved; Overview and widget changes are absent from the introduced delta. Both retained rank-up requests are addressed by the revised dashboard-only body and new proof. Supplied checks report nine successes; this read-only review did not execute tests. (Tests/CodexBarTests/DashboardReportedCostTests.swift:9, a8ee4c977986)
• Related work is distinct: The supplied GitHub record verifies Add DeepSeek and OpenRouter widgets #3743 merged widget selection and credit-balance support, not this dashboard spend projection. Open Keep retained models on the shared usage card #3747 concerns retained model rankings in share cards and does not replace this work.

Likely related people:

• steipete: Suggested for follow-up; no historical authorship or introduction is verified. (role: unverified routing candidate; confidence: low)
• unknown: The claimed source-line change could not be verified from bounded local history. (role: source history unknown; confidence: low)

Rating scale

Score	Internal tier	Crab rank	Meaning
6/6	S	🦀 challenger crab	Exceptional readiness
5/6	A	🦞 diamond lobster	Very strong readiness
4/6	B	🐚 platinum hermit	Good normal PR; ordinary maintainer review
3/6	C	🦐 gold shrimp	Useful, but confidence is limited
2/6	D	🦪 silver shellfish	Proof or implementation needs work
1/6	F	🧂 unranked krab	Not merge-ready
N/A	NA	🌊 off-meta tidepool	Rating does not apply

Overall follows the weaker of proof and patch quality.
Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

Workflow

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

History

Review history (3 earlier review cycles)

• reviewed 2026-09-18T21:31:03.289Z sha 8fcf687 :: needs real behavior proof before merge. :: [P2] Preserve UTC bucket semantics before deriving today's spend
• reviewed 2026-09-18T21:40:22.967Z sha 010df21 :: needs real behavior proof before merge. :: [P2] Preserve UTC semantics before deriving today's spend | [P2] Exclude stale publication inputs before building Overview | [P1] Honor hidden sources in the ownership-mismatch fallback
• reviewed 2026-09-20T02:34:17.406Z sha a8ee4c9 :: needs real behavior proof before merge. :: none

[Chipagosfinest]

Extended past the CLI dashboard: Overview now merges live costUsage when the spend publication omitted OpenRouter, and widgets persist Activity 30-day cost/tokens. OpenRouter is selectable in the widget gallery.

Overlap with #3743: that PR shows credit balance; this one shows reported Activity spend on Overview and widget cost metrics.

@clawsweeper re-review

[clawsweeper]

🦞🧹
ClawSweeper re-review requested.

I asked ClawSweeper to review this item again.
Action: item re-review queued (workflow sweep.yml, event exact_review_queue).
Result: when the review finishes, ClawSweeper will create the durable review comment if needed or update the existing comment in place.