The latest articles on DEV Community by RisingInsight (@risinginsight). https://dev.to/risinginsight https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4232023%2F3c2b3bbd-7ee9-47fe-9ed9-ba562ea0f9a9.jpg https://dev.to/risinginsight en RisingInsight Sat, 07 Nov 2026 11:56:04 +0000 https://dev.to/risinginsight/starport-one-local-endpoint-for-claude-code-codex-cli-and-cline-with-credential-pools-and-auto-kc2 https://dev.to/risinginsight/starport-one-local-endpoint-for-claude-code-codex-cli-and-cline-with-credential-pools-and-auto-kc2 <p>I kept re-wiring configs between Codex CLI (Responses API), Claude Code (Anthropic Messages) and Cline/SDKs (Chat Completions), while my model access was scattered across several platforms — each with its own login, quota and credential format. So I wrote a gateway that collapses both layers.</p> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fcnbml02lass1tx497y84.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fcnbml02lass1tx497y84.png" alt="Accounts center" width="800" height="450"></a></p> <h2> One endpoint, three protocols </h2> <p>Three inbound faces — <code>/v1/chat/completions</code>, <code>/v1/messages</code> (Anthropic), <code>/v1/responses</code> (Codex CLI) — fold into one internal wire, then re-encode into four outbound dialects (openai / anthropic / gemini / responses). Codex CLI, Claude Code and Cline can all point at one local endpoint; switching backends is a one-line model-name change.</p> <h2> Credential pools with auto check-ins </h2> <p>Multi-platform accounts live in one pool: cost-tiered selection (measured, not list price), fast-expiring-credit weighting, 429 exponential backoff, hard cooldown on 402, breaker, and per-conversation stickiness to keep the prompt cache warm.</p> <p>The part most gateways skip: <strong>scheduled housekeeping</strong>. Daily check-ins, activity reports and token keep-alive run per-day idempotently and self-heal after restarts — five days of soak testing with zero missed check-ins across reboots.</p> <p>And the credit data feeds back into routing: fast-expiring credits get weighted first, per-model costs are measured (not list price), and a credit floor stops paid models from draining accounts.</p> <p><a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fhf5ofd9xopiyhjtyawg8.png" class="article-body-image-wrapper"><img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2Fhf5ofd9xopiyhjtyawg8.png" alt="Usage" width="800" height="450"></a></p> <h2> Web panel </h2> <p>Accounts, usage aggregation, task center, API keys with per-day/week/month token budgets, request logs with zero credential plaintext.</p> <p>Single Go binary (Win/Linux/macOS) or Docker. Listens on 127.0.0.1 by default. No telemetry. MIT.</p> <p>Repo: <a href="https://github.com/RisingInsight/startport" rel="noopener noreferrer">https://github.com/RisingInsight/startport</a> — issues welcome!</p> ai go opensource llm