fix(usage): align cost with usage

[ymkiux]

Fix usage token total fallback so estimated cost matches usage.

Summary by CodeRabbit

• Bug Fixes

  • Improved session token total calculations when explicit total is missing; now sums individual token breakdowns (input, cached input, output, reasoning output) for accurate usage metrics.

• Tests

  • Added unit tests for token aggregation and session model extraction logic.

[ymkiux]

说明：此仓库 main 为保护分支，禁止直接 push，且合并需要至少 1 个非作者的 Approve。我无法用同一身份给自己开的 PR 点 Approve，因此无法合并该 PR；在 main 未合入“连接历史”的提交前，GitHub 会一直拒绝 reopen #95（提示 head 与 main 无共同历史）。若你让另一位 maintainer approve 并 merge #97，再执行 reopen #95 即可。

[coderabbitai]

Warning

Rate limit exceeded

@ymkiux has exceeded the limit for the number of commits that can be reviewed per hour. Please wait 22 minutes and 4 seconds before requesting another review.

Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 22 minutes and 4 seconds.

⌛ How to resolve this issue?

After the wait time has elapsed, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout.

Please see our FAQ for further information.

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: e629e0ee-5538-4dbe-9810-77f7c5d527d3

📥 Commits

Reviewing files that changed from the base of the PR and between f055cdd and b28f215.

📒 Files selected for processing (5)

• cli/session-usage.concurrent.js
• cli/session-usage.js
• cli/session-usage.models.js
• web-ui/partials/index/panel-usage.html
• web-ui/styles/sessions-usage.css

📝 Walkthrough

Walkthrough

Added robust session token aggregation and a new CLI session-usage core: a helper computes total tokens from breakdown fields when explicit totals are missing; CLI now delegates session usage listing to a new core module that normalizes model names (including file scanning) with caching and concurrency limits. Tests updated/added.

Changes

Cohort / File(s)	Summary
Unit Tests tests/unit/web-ui-logic.test.mjs	Added test verifying buildUsageChartGroups computes totalTokens from token breakdown fields when totalTokens is absent.
Token Calculation Logic web-ui/logic.sessions.mjs	Added readSessionTotalTokens(session) for numeric validation and fallback aggregation; buildUsageChartGroups now uses it instead of direct parsing.
CLI delegation cli.js	listSessionUsage now delegates implementation to listSessionUsageCore from ./cli/session-usage and passes required deps; removed inline aggregation logic.
Session usage core cli/session-usage.js	New module implementing listSessionUsageCore, concurrency limiter, model-name normalization, JSONL model extraction, per-file caching/read limiter, and session normalization flow (browse → file scan → summary parse).
Backend tests tests/unit/session-usage-backend.test.mjs	Updated tests to import/use listSessionUsageCore; adjusted expectations for parsing/scan behavior and ordering assertions.

Sequence Diagram(s)

sequenceDiagram
participant CLI as "cli.js (caller)"
participant Core as "listSessionUsageCore"
participant Browse as "listSessionBrowse"
participant FileReader as "createSessionModelsFileReader"
participant Parsers as "parseCodex/parseClaude"
CLI->>Core: listSessionUsageCore(params, deps)
Core->>Browse: request session browse (bounded limit)
Browse-->>Core: session list
Core->>Core: normalize per-session fields (models/modelName)
alt models present in browse
Core-->>CLI: include normalized session
else models missing
Core->>FileReader: read models from session file (concurrency-limited, cached)
FileReader-->>Core: models[] or []
alt models found
Core-->>CLI: include normalized session
else
Core->>Parsers: parse limited-bytes session summary
Parsers-->>Core: summary (models/modelName/provider)
Core-->>CLI: include normalized session (if obtained)
end
end

Loading

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

Possibly related PRs

• fix: stabilize web ui startup and reduce usage load #83 — Modifies buildUsageChartGroups and tests around totalTokens aggregation; closely related to the token-fallback changes.
• feat: add sessions usage tab #74 — Introduced buildUsageChartGroups in the same file; directly related to subsequent aggregation adjustments.
• refactor: split web ui modules and harden provider switching #68 — Prior splitting/changes to logic.sessions.mjs that interact with the current token-aggregation updates.

Suggested reviewers

• awsl233777
• SurviveM

Poem

🐰
I hop through sessions, tiny and spry,
When totals vanish I give it a try.
Sum inputs, outputs, reasoning too,
Charts sing true — now numbers are new! 🥕📊

🚥 Pre-merge checks | ✅ 1 | ❌ 2

❌ Failed checks (1 warning, 1 inconclusive)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.
Title check	❓ Inconclusive	The title 'fix(usage): align cost with usage' is vague and does not clearly describe the main technical changes in the PR. The changes involve adding fallback logic for computing session token totals from breakdown fields, delegating session-usage logic to a new module, and refactoring related code. The title does not meaningfully convey what problem is being fixed or what the changeset accomplishes.	Consider a more descriptive title such as 'fix(usage): compute missing session totalTokens from token breakdown' or 'fix(usage): add fallback token total calculation from breakdown fields' to clarify the technical intent.

✅ Passed checks (1 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch chore/link-pr95-history

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@cli/session-usage.js`:
- Around line 1-21: The limiter can let new tasks skip the queue because
`active` is decremented before a queued waiter actually acquires the slot; fix
`createConcurrencyLimiter` so slot acquisition is atomic by moving the increment
into the wake logic: change `next()` to loop while there are queued waiters and
`active < max`, and for each picked waiter increment `active` before calling its
resolve (so the waiter resumes already owning a slot); keep the `active -= 1` in
the `finally` of the worker, and ensure the immediate-path still increments
`active` before running the task. This guarantees queued waiters increment
`active` inside `next()` and cannot be raced past by concurrently submitted
tasks.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 07a2aa58-16ff-462b-9172-7f8650bc883d

📥 Commits

Reviewing files that changed from the base of the PR and between c8cfae5 and f055cdd.

📒 Files selected for processing (5)

• cli.js
• cli/session-usage.js
• tests/unit/session-usage-backend.test.mjs
• tests/unit/web-ui-logic.test.mjs
• web-ui/logic.sessions.mjs

✅ Files skipped from review due to trivial changes (1)

• tests/unit/web-ui-logic.test.mjs

🚧 Files skipped from review as they are similar to previous changes (1)

• web-ui/logic.sessions.mjs

📜 Review details

🧰 Additional context used

🧠 Learnings (1)

📓 Common learnings

Learnt from: github-actions[bot]
Repo: SakuraByteCore/codexmate PR: 0
File: :0-0
Timestamp: 2026-04-04T01:36:26.964Z
Learning: In PR `#68` for SakuraByteCore/codexmate, github-actions[bot] again flagged that CodeRabbit was introducing breaking changes. The maintainer's absolute constraint (reinforced across PRs `#55`, `#59`, `#67`, `#68`) is: no breaking changes in any review suggestion; limit all feedback to correctness, safety, and quality only within the existing design, without refactoring or altering APIs, interfaces, or behavior.

🔇 Additional comments (5)

cli.js (2)

141-141: Import wiring looks good.

The new core import follows the existing CommonJS style and keeps the session-usage implementation isolated from the CLI entrypoint.

---

4263-4272: Delegation preserves the existing API.

listSessionUsage(params = {}) keeps its public signature while passing the existing dependencies and limits into the core module, which matches the PR’s goal without changing callers.

tests/unit/session-usage-backend.test.mjs (3)

56-64: LGTM on the binding injection.

Defaulting listSessionUsageCore from the loaded usageCore while still letting callers override via bindings keeps existing tests decoupled from the new core module. The bindings || {} spread also safely handles nullish arguments despite the = {} default.

---

353-356: Order-insensitive assertion is appropriate.

Switching to sorted-array comparison correctly reflects that listSessionUsageCore now fans out summary parsing through a concurrency limiter (Promise.all over limitNormalize, 32 concurrent tasks in cli/session-usage.js), so call order is no longer deterministic. Good defensive fix to avoid flaky tests.

---

359-423: Test aligns with new 3-tier fallback.

The renamed test correctly exercises the early-return branch in listSessionUsageCore where baseModels derived from the lightweight entry's model field short-circuits before any file I/O or summary parsing. Asserting codexParses is empty and models === ['gpt-5.3-codex'] properly validates that the full-file scan and parseCodexSessionSummary are both skipped. The >64KB fixture size check also guards the intent of the test.

[awsl233777]

Required checks are green, and the usage/cost alignment change looks ready to merge.