codex-lb is a proxy and load balancer that combines multiple ChatGPT accounts behind one service, tracks their usage, manages API keys, and provides a dashboard. It is for people using Codex CLI, OpenCode, or other OpenAI-compatible clients who need account pooling and usage monitoring. Catalogue add-ons can operate it through its compatible endpoints.
Borrowing it
Nothing to install: this file belongs to Soju06/codex-lb. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Soju06/codex-lb/main/.agents/skills/codex-review-loop/SKILL.mdgit clone --depth 1 https://github.com/Soju06/codex-lbWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/soju06/codex-lb/codex-review-loop)<a href="https://agentmods.dev/skills/soju06/codex-lb/codex-review-loop"><img src="https://agentmods.dev/badge/skills/soju06/codex-lb/codex-review-loop/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/soju06/codex-lb/codex-review-loop"><img src="https://agentmods.dev/badge/skills/soju06/codex-lb/codex-review-loop.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium MCP Rug Pull · line 149 uvx/uv tool run commands without ==version create a rug-pull risk.Fix: Pin the version: uvx package-name==1.2.3
- medium MCP Rug Pull · line 150 uvx/uv tool run commands without ==version create a rug-pull risk.Fix: Pin the version: uvx package-name==1.2.3
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.01705 |
| Opus 5.5 | $0.00014 | $0.00682 |
| Sonnet 5.5 | $0.00007 | $0.00341 |
| Haiku 4.5 | $0.00004 | $0.00170 |
Grade A, and why
codex-review-loop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 195 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Codex Review Loop
Adversarial code review via Codex CLI, structured finding analysis, HITL-gated fix loop, optional re-verification.
Arguments
--pr <N>— review PR number N (default: auto-detect most recent open PR)--base <branch>— review changes against base branch--uncommitted— review uncommitted local changes
Phases
- Scope Resolution — determine review target (PR/branch/uncommitted)
- Codex Review — launch adversarial review via Codex CLI (~20-50 min)
- Finding Analysis — parse raw output into structured findings
- Atomic Fix Loop — fix, verify, commit each approved finding
- Re-review Loop — re-run review until 0 findings (max 3 iterations)
- Final Report — summary of all findings, fixes, and verification status
Phase 1: Scope Resolution
Determine the review target and extract the base branch.
- PR mode (default): Use
gh pr view/gh pr listto resolve base branch and display PR metadata (title, file count, additions/deletions). - Branch mode (
--base): Use specified base branch directly. - Uncommitted mode (
--uncommitted): Review working tree changes.
If no argument is given, auto-detect the user's most recent open PR and confirm.
Phase 2: Codex Review
Launch the adversarial review as a background process.
- Run the review script with
run_in_background=true:
Note: Codex CLI v0.105.0+ does not support combiningbash <skill-dir>/scripts/codex-subagent.sh --base <branch>--base/--commitwith a custom prompt. The built-in review logic is used automatically. For--uncommittedmode (no diff target), pipe stdin for custom instructions:cat <prompt-file> | bash <skill-dir>/scripts/codex-subagent.sh --uncommitted - Inform the user the review is running (~20-50 min).
- The script parses Codex output and returns the final review text. Review
rollouts intentionally remain persistent. If terminal output is lost, use
codex resume --include-non-interactiveto locate the review, orcodex resume <SESSION_ID>when its ID is known. Do not add--ephemeralto the wrapper. The wrapper also relies on the configured non-interactive approval/sandbox policy because Codex CLI 0.147.0 removed the historical--full-autoargument fromexec review.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 195 lines · 36 tokens per session scan A 10205bf38a40
codex-review-loop is a skill published in the GitHub repository Soju06/codex-lb (3,382 stars, last pushed 2d ago), licensed MIT. It adds 36 tokens to every session and 1,705 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
code-review
Review the current diff, or a PR number/branch/path target, for correctness bugs (plus reuse/simplification/efficiency cleanups where the model's review recipe covers them) at the given effort level (low|medium|high|xhigh|max: from few high-confidence findings up to many, some of them uncertain; ultra: deep…
github
GitHub via gh CLI: PRs, issues, reviews, repos, auth.
gh-issues
Fetch GitHub issues, spawn sub-agents to implement fixes and open PRs, then monitor and address PR review comments. Usage: /gh-issues [owner/repo] [--label bug] [--limit 5] [--milestone v1.0] [--assignee @me] [--fork user/repo] [--watch] [--interval 5] [--reviews-only] [--cron] [--dry-run] [--model glm-5]…
github
GitHub operations via gh CLI: issues, PRs, CI runs, code review, API queries. Use when: (1) checking PR status or CI, (2) creating/commenting on issues, (3) listing/filtering PRs or issues, (4) viewing run logs. NOT for: complex web UI interactions requiring manual browser flows (use browser tooling when available)…
pr-creator
Create well-structured pull requests with proper descriptions and labels.
github_read
Read access: repos, issues, PRs, files, branches, workflows, runs.