ci: swap direct-OpenAI review slot from gpt-5 to gpt-5.6-luna (cheaper, newer)

[seonghobae]

Why

The opencode-review model pool is currently exhausting org-wide (e.g. html4tree PR #166, failed twice):

• github-models/*: "This account has reached its budget limit" + 4000-token request cap on the starved tier
• openai/gpt-5 (direct): insufficient_quota — and legacy gpt-5 is also the pool's most expensive direct slot at $1.25/M input, $10/M output

What

Replace the direct-OpenAI candidate with GPT-5.6 Luna — the newest model family's cost-efficient tier, per OpenAI's official pricing:

	gpt-5 (legacy)	gpt-5.6-luna
Input /1M	$1.25	$1.00
Output /1M	$10.00	$6.00
Context	400K	1M
Generation	Aug 2025	newest (5.6 family)

• Registers gpt-5.6-luna in the direct-OpenAI provider catalog (reasoning effort high, same shape as the existing gpt-5 entry, which stays for other call sites)
• Points the OPENCODE_MODEL_CANDIDATES direct-OpenAI slot at it
• Updates the pinned contract-test expectations

Luna is a full-size model, so it does not trip the pool's is_low_sensitivity_candidate mini/nano exclusion.

Verification

• python -m pytest tests/ — 375 passed, 8 skipped

Note

This lowers per-review cost but the OpenAI account still needs its quota/billing topped up before the direct slot produces verdicts again; the GitHub Models budget cap is likewise a billing setting, not a code issue.

🤖 Generated with Claude Code

[opencode-agent]

OpenCode Review Overview

• Head SHA: 83f95ca00b8db124e1ba58c25f62c6680d08be60
• Workflow run: 29223952170
• Workflow attempt: 2
• Gate result: APPROVE (exit 0)

Changed-File Evidence Map

flowchart LR
  PR["PR changed files"] --> Evidence["OpenCode bounded evidence"]
  Evidence --> S1["Workflow: opencode-review.yml"]
  S1 --> I1["GitHub Actions review job"]
  I1 --> R1["Review risk: Workflow: opencode-review.yml"]
  R1 --> V1["actionlint plus required checks"]
  Evidence --> S2["Test: test_opencode_agent_contract.py"]
  S2 --> I2["regression suite"]
  I2 --> R2["Review risk: Test: test_opencode_agent_contract.py"]
  R2 --> V2["targeted test run"]

Loading

[opencode-agent]

Pull request overview

Approval sufficiency: bounded evidence supplied affirmative approval evidence for changed files, coverage/docstring posture, risk surfaces, and current-head verification; approval is not based merely on the absence of known blockers.
Verification posture: CodeGraph evidence was initialized and bounded current-head evidence reviewed for changed-file evidence including .github/workflows/opencode-review.yml, tests/test_opencode_agent_contract.py.
Linter/static: workflow/static review evidence is bounded by the current-head GitHub Checks gate and changed-file evidence.
TDD/regression: coverage execution evidence and focused changed hunks were reviewed from bounded-review-evidence.md.
Coverage: coverage execution evidence reports supported repository test suites passed.
Docstring coverage: coverage execution evidence reports configured repository docstring gates passed or docstring coverage was advisory.
DAG: CodeGraph/source-backed behavior map connects .github/workflows/opencode-review.yml to the affected review, runtime, or workflow path and required checks.
PoC/execution: coverage-evidence job executed on the current head and reported PASS.
DDD/domain: workflow and repository-governance invariants were reviewed against changed files in bounded evidence.
CDD/context: CodeGraph evidence, changed-file history, and focused hunks were reviewed from bounded-review-evidence.md.
Similar issues: changed-file history evidence was reviewed for comparable local precedents.
Claim/concept check: bounded evidence, repository source, current-head workflow evidence, and, where numeric, scientific, statistical, or literature-backed claims are affected, original-paper/formula evidence and parameter-recovery expectations were used for claims.
Standards search: standards and external-source checks are delegated to configured OpenCode web_search/Context7/DeepWiki sources when applicable; no evidence-backed standards blocker is present in bounded evidence.
Compatibility/convention: changed workflow/script conventions, object naming, and reserved-word safety for schema/API/config/code surfaces were checked in bounded evidence.
Breaking-change/backcompat: deployment evidence and changed-file history were checked for backward-compatibility risk.
Performance: changed surfaces were checked for performance risk in bounded evidence.
Developer experience: changed automation, review, test, setup, and maintenance surfaces were checked for helpful or obstructive DX impact in bounded evidence.
User experience: connected user, operator, API, CLI, documentation, review-comment, status-check, rendering, and workflow-reader behavior was checked for contradictions against code, docs, and tests in bounded evidence.
Visual/DOM: Playwright visual, DOM locator, ARIA snapshot, console, and responsive evidence were checked when a web UI surface was present; for non-web surfaces, API/CLI/log/docs/workflow interaction evidence was reviewed instead.
Accessibility/i18n: accessibility, localization, and human-readable text surfaces were checked where UI, CLI, API message, docs, logs, or review text changed.
Supply-chain/license: dependency, package, model, container, and external-tool changes were checked in bounded evidence.
Packaging: package, build, test, lint, and security contracts were checked in bounded evidence.
Security/privacy: workflow-token, review-gate, and repository-automation security/privacy boundaries were checked in bounded evidence.

Findings

No blocking findings.

Adversarial validation

{"status":"passed","probes":[{"path":".github/workflows/opencode-review.yml","line":2998,"hypothesis":"openai/gpt-5.6-luna is not actually defined in the provider models, causing runtime selection failure when chosen by the pool.","attack_or_counterexample":"Read provider.openai.models in the workflow to confirm presence and reasoning options for gpt-5.6-luna.","evidence":"Found entry at lines ~2998–3014: \"gpt-5.6-luna\" with tool_call:true, reasoning:true, options.reasoningEffort:\"high\", variants.high.reasoningEffort:\"high\", and limit set.","outcome":"falsified"},{"path":".github/workflows/opencode-review.yml","line":3277,"hypothesis":"OPENCODE_MODEL_CANDIDATES still references openai/gpt-5, causing mismatch with tests and provider config.","attack_or_counterexample":"Inspect the workflow env for OPENCODE_MODEL_CANDIDATES.","evidence":"Line ~3277 includes: OPENCODE_MODEL_CANDIDATES: \"github-models/deepseek/deepseek-v3-0324 openai/gpt-5.6-luna github-models/openai/gpt-5 ...\"","outcome":"falsified"}],"residual_risk":"Low. If OpenAI quota or pricing characteristics for gpt-5.6-luna shift or the model is temporarily unavailable, the pool will fall back to established GitHub Models candidates. No new secret handling or permission surfaces introduced."}

Evidence

• Result: APPROVE
• Reason: Swaps the direct-OpenAI candidate from gpt-5 to gpt-5.6-luna and updates tests accordingly; config and tests are consistent, coverage is green, and no security or workflow contract regressions were found.
• Scope: central OpenCode/Strix review-process
• Changed files: 2
• Head SHA: 83f95ca00b8db124e1ba58c25f62c6680d08be60
• Workflow run: 29223952170
• Workflow attempt: 2

This approval path is limited to ContextualWisdomLab/.github central review-process self-repair.