test(opencode): Ollama OpenAI-compat baseURL /v1 residual

[TimothyVang]

Summary

• Confirms bash scripts/check-httpapi-error-contracts.sh passes (26 tests).
• Caseforge agent 404 page not found is not an opencode provider/OpenAI path bug: @ai-sdk/openai-compatible POSTs ${baseURL}/chat/completions; Ollama only serves that under /v1. Bare VERDICT_LLM_BASEURL=http://host:11434 hits /chat/completions → plain page-not-found. Normalization belongs in caseforge (engine lane).
• Residual: offline provider tests lock that this runtime preserves baseURL as configured (does not auto-append /v1); optional smoke scripts/check-ollama-openai-compat.sh; short note in VERDICT-FORK.md.

Evidence

bash scripts/check-httpapi-error-contracts.sh  # 26 pass
bun test test/provider/provider.test.ts -t "Ollama-style"  # 2 pass
VERDICT_LLM_BASEURL=http://10.126.60.100:11434/v1 bash scripts/check-ollama-openai-compat.sh
# PASS: Ollama OpenAI-compat reachable (v1 OpenAI JSON error vs bare page-not-found)

Test plan

• Error-contract script
• New Ollama-style provider tests
• Optional live Ollama smoke against Spark
• No TUI rebrand / no engine behavior change

[github-actions]

Thanks for your contribution!

This PR doesn't have a linked issue. All PRs must reference an existing issue.

Please:

1. Open an issue describing the bug/feature (if one doesn't exist)
2. Add Fixes #<number> or Closes #<number> to this PR description

See CONTRIBUTING.md for details.

[github-actions]

This PR doesn't fully meet our contributing guidelines and PR template.

What needs to be fixed:

• PR description is missing required template sections. Please use the PR template.

Please edit this PR description to address the above within 2 hours, or it will be automatically closed.

If you believe this was flagged incorrectly, please let a maintainer know.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 88cc2e505e

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Reject bare VERDICT_LLM_BASEURL in smoke test

When VERDICT_LLM_BASEURL is set to the misconfigured bare Ollama root (http://host:11434), this script strips/normalizes it into origin and then probes ${origin}/v1/chat/completions, so it reports PASS even though the runtime would still POST to ${VERDICT_LLM_BASEURL}/chat/completions and hit the documented 404 page not found. Since this smoke check is meant to catch that exact missing-/v1 configuration, validate the supplied base URL (or probe the raw configured request path) before constructing the corrected /v1 URL.

Useful? React with 👍 / 👎.