skills.rest
zero-yx
zero-yxCOMMUNITY
Explore all skills by this author →
GitHub

approving-llm-gateway-account-batches

Automates batch preflight, validation, issuance, and patching of pending LLM Gateway account-contribution requests via admin APIs.

2|Updated Jun 18, 2026
Search Tags:#automation#codex#admin api#account provisioning#llm gateway#batch approval#proxy assignment
One-click install
npx skills add https://github.com/zero-yx/static_flow --skill approving-llm-gateway-account-batches-zero-yx
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: approving-llm-gateway-account-batches
Source: https://github.com/zero-yx/static_flow/tree/main/skills/approving-llm-gateway-account-batches
Command: npx skills add https://github.com/zero-yx/static_flow --skill approving-llm-gateway-account-batches-zero-yx

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Processing queued LLM Gateway account-contribution requests one by one through admin APIs is slow and error-prone, especially when each account needs credential preflight checks, deterministic proxy assignment, and post-issue configuration. ## Core Features & Use Cases - Batch Admin Workflow: Fetches pending requests, preflights Codex credentials (OAuth refresh plus models probe through a local proxy), validates, approves, and issues accounts in bulk. - Deterministic Proxy Assignment: Selects the least-used active proxy config with name-based tie-breaking so one batch is balanced across proxies. - Standard Post-Issue Patching: Applies auto_refresh_enabled, fixed proxy mode, concurrency limits, and randomized start intervals, then attempts a best-effort usage refresh. - Use Case: After a public batch submit creates 20 pending requests with prefix pickup7_, run the helper script in dry-run mode to review the plan, then apply it to validate, issue, and configure every account end to end. ## Quick Start Ask the agent to run the approval script against the local admin API at http://127.0.0.1:19182 with your account prefix and expected count, first as a dry run and then with --apply to issue and patch the accounts.

Frequently Asked Questions about approving-llm-gateway-account-batches

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I batch approve pending LLM Gateway account requests?▼

Run the approve_account_contribution_batch.py script with --admin-base-url, --account-prefix, and --expected-count. Start with a dry run to review the plan, then rerun with --apply to validate, issue, patch, and usage-refresh each account.

How does the script preflight Codex account credentials?▼

It attempts an OAuth refresh against auth.openai.com and probes the Codex models endpoint through the local proxy at http://127.0.0.1:11116. Accounts are issued only when the models probe succeeds, even if refresh fails.

What happens if token refresh fails but models access works?▼

The account is still validated and issued, but it is patched with auto_refresh_enabled set to false. This supports auth bundles with working access tokens but broken or short-lived refresh credentials.

How are proxies assigned to imported accounts?▼

Only active proxy configs are considered. The script counts current usage across active Codex accounts, picks the least-used proxy, breaks ties by proxy name, and updates the in-memory count so one batch stays balanced.

Does a failed usage refresh roll back an issued account?▼

No. Post-issue usage refresh is best-effort with configurable retries. Failures are recorded in the result JSON under /tmp/ without rolling back the issued account or failing the whole batch item.

What are the limitations of this batch approval workflow?▼

It requires pending requests to already exist, an exact expected count match, at least one active proxy config, and a reachable admin API. It only sequences existing admin endpoints and does not bypass the backend review flow.

Related Skills

Explore More Software Engineering Skills→
NethereumNethereum
2.3k

account-abstraction

Implements ERC-4337 Account Abstraction with Nethereum to send UserOperations through bundlers in .NET.

Official
Intermediate
fryanpanfryanpan

token-watch

Monitors Claude account token quotas across multiple pools and reports estate status with rotation verdicts.

Community
Advanced
LXE123LXE123
1

fba-shipment-wms-box-download

Download WMS packing Excel data for Amazon FBA shipments via CLI.

Official
Intermediate
polymerdaopolymerdao
1

polymer-pay-twit

Retrieve public X/Twitter data through the Polymer Pay proxy.

Official
Intermediate