Skip to content

fix(opencode-review): native OpenAI backend as lead review model (unblock org-wide merge freeze) #350

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
seonghobae wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix(opencode-review): native OpenAI backend as lead review model (unblock org-wide merge freeze) #350

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
80 changes: 69 additions & 11 deletions .github/workflows/opencode-review.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2006,7 +2006,7 @@ jobs:
"$schema": "https://opencode.ai/config.json",
"model": "github-models/deepseek/deepseek-r1-0528",
"small_model": "github-models/deepseek/deepseek-v3-0324",
"enabled_providers": ["github-models"],
"enabled_providers": ["openai", "github-models"],
"lsp": true,
"mcp": {
"codegraph": {
Expand Down Expand Up @@ -2132,6 +2132,50 @@ jobs:
}
},
"provider": {
"openai": {
"npm": "@ai-sdk/openai",
"name": "OpenAI (direct)",
"options": {
"baseURL": "https://api.openai.com/v1",
"apiKey": "{env:OPENAI_API_KEY}"
},
"models": {
"gpt-5": {
"name": "OpenAI GPT-5 (direct)",
"tool_call": true,
"reasoning": true,
"options": {
"reasoningEffort": "high"
},
"variants": {
"high": {
"reasoningEffort": "high"
}
},
"limit": {
"context": 400000,
"output": 128000
}
},
"gpt-5-mini": {
"name": "OpenAI GPT-5 Mini (direct)",
"tool_call": true,
"reasoning": true,
"options": {
"reasoningEffort": "high"
},
"variants": {
"high": {
"reasoningEffort": "high"
}
},
"limit": {
"context": 400000,
"output": 128000
}
}
}
},
"github-models": {
"npm": "@ai-sdk/openai-compatible",
"name": "GitHub Models",
Expand Down Expand Up @@ -2339,20 +2383,31 @@ jobs:
env:
STRIX_GITHUB_MODELS_TOKEN: ${{ secrets.STRIX_GITHUB_MODELS_TOKEN || github.token }}
GITHUB_TOKEN: ${{ secrets.STRIX_GITHUB_MODELS_TOKEN || github.token }}
# Native OpenAI backend for the lead review model. GitHub Models
# rate-limits every request and caps bodies at ~4000 tokens, so the
# rate-starved shared pool never returned a verdict; hitting
# api.openai.com directly with the org OPENAI_API_KEY gives the lead
# model a working, un-throttled backend. Resolves {env:OPENAI_API_KEY}
# in the opencode.jsonc "openai" provider block.
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
USE_GITHUB_TOKEN: "true"
SHARE: "false"
NPM_CONFIG_IGNORE_SCRIPTS: "true"
NO_COLOR: "1"
# Ordered contract-reliability first, then quota, with the rate-starved
# flagships last. The mini reasoning models (o4-mini, o3-mini, gpt-5-
# mini/nano/chat) reliably emit the strict review contract — every
# required label and only source-backed findings — so they lead. The
# high-quota non-reasoning models (deepseek-v3, mistral, llama-4) emit
# bare or hallucinated reviews the publish/approve gates reject, so
# they are fallbacks only. gpt-5/o3 ("Reasoning" tier, 8-12 req/day)
# stay last: first-placing them stalled every review until timeout
# because a rate-limited/hung flagship never fell back.
OPENCODE_MODEL_CANDIDATES: "github-models/openai/o4-mini github-models/openai/o3-mini github-models/openai/gpt-5-mini github-models/openai/gpt-5-nano github-models/openai/gpt-5-chat github-models/deepseek/deepseek-r1-0528 github-models/deepseek/deepseek-r1 github-models/deepseek/deepseek-v3-0324 github-models/mistral-ai/mistral-medium-2505 github-models/meta/llama-4-maverick-17b-128e-instruct-fp8 github-models/meta/llama-4-scout-17b-16e-instruct github-models/openai/o3 github-models/openai/gpt-5"
# Lead with the NATIVE OpenAI backend (openai/gpt-5, openai/gpt-5-mini
# via api.openai.com with the org OPENAI_API_KEY). GitHub Models
# rate-limited ("Too many requests") and 4000-token-capped
# (413 tokens_limit_reached) EVERY model in the shared pool, so the
# reviewer never produced a verdict and every run hung to the 350-min
# timeout — a 100% org-wide failure. The native provider is not subject
# to those limits, so it can actually complete and approve. The
# existing github-models entries stay as fallbacks (tried only if the
# native key is missing or the direct call fails).
# github-models ordering rationale (unchanged): contract-reliable mini
# reasoning models first, high-quota non-reasoning models next, and the
# rate-starved github-models flagships (gpt-5/o3, 8-12 req/day) last so
# a throttled/hung leader always falls back instead of eating the step.
OPENCODE_MODEL_CANDIDATES: "openai/gpt-5 openai/gpt-5-mini github-models/openai/o4-mini github-models/openai/o3-mini github-models/openai/gpt-5-mini github-models/openai/gpt-5-nano github-models/openai/gpt-5-chat github-models/deepseek/deepseek-r1-0528 github-models/deepseek/deepseek-r1 github-models/deepseek/deepseek-v3-0324 github-models/mistral-ai/mistral-medium-2505 github-models/meta/llama-4-maverick-17b-128e-instruct-fp8 github-models/meta/llama-4-scout-17b-16e-instruct github-models/openai/o3 github-models/openai/gpt-5"
# One attempt per model, then fall through to the next model. Retrying
# the SAME model 5x let a rate-limited/hung leader consume the whole
# step, so the pool never reached a healthy fallback model.
Expand Down Expand Up @@ -2732,6 +2787,9 @@ jobs:
CHECK_LOOKUP_GH_TOKEN: ${{ github.token }}
GH_REPOSITORY: ${{ github.event.pull_request.base.repo.full_name || github.event.inputs.target_repository || github.repository }}
STRIX_GITHUB_MODELS_TOKEN: ${{ secrets.STRIX_GITHUB_MODELS_TOKEN || github.token }}
# Exposed so the "openai" provider in opencode.jsonc resolves during the
# failed-check diagnosis opencode run that shares this config.
OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
OPENCODE_APP_TOKEN: ${{ steps.opencode_app_token.outputs.token }}
OPENCODE_EVIDENCE_FILE: ${{ runner.temp }}/opencode-review-evidence.md
OPENCODE_FAILED_CHECK_EVIDENCE_FILE: ${{ runner.temp }}/opencode-failed-check-evidence.md
Expand Down
Loading