[WIP] Fix failed jobs in Smoke Goose workflow

[Copilot]

• Investigate check_token_telemetry failure in Smoke Goose run 30944874071
• Root cause: GOOSE_PROVIDER__HOST / GOOSE_PROVIDER__API_KEY / GOOSE_PROVIDER__TYPE are Goose desktop (Electron) settings and are ignored by the Goose CLI. The Rust OpenAI provider resolves its endpoint from OPENAI_HOST (highest priority) then OPENAI_BASE_URL, so Goose sent the Copilot token to the OpenAI proxy port 10000 and got 401 Unauthorized → no LLM calls → empty token-usage.jsonl → check_token_telemetry failed
• Replace the ignored GOOSE_PROVIDER__* vars with OPENAI_HOST (AWF Copilot api-proxy at 172.30.0.30:10002) and OPENAI_BASE_PATH in .github/workflows/shared/goose.md
• Recompile workflows (make recompile)
• Run make agent-report-progress — lint clean, lock files in sync, tests pass

• Fixes [aw] Failed jobs: Smoke Goose #50383

---

Generated by 👨‍🍳 PR Sous Chef · gpt54 · 8.54 AIC · ⌖ 6.21 AIC · ⊞ 8.3K · ◷
_{Comment /souschef to run again}

[pelikhan]

/smoke-goose

[github-actions]

🪿 Smoke Goose Goose delivered.

[pelikhan]

@copilot recompile and review this error:

Warning: Failed to start extension 'safeoutputs' (failed to initialize MCP client: Send message error Transport [rmcp::transport::worker::WorkerTransport<rmcp::transport::streamable_http_client::StreamableHttpClientWorker<reqwest::async_impl::client::Client>>] error: Client error: error sending request for url (http://localhost:8080/mcp/safeoutputs), when send initialize request), continuing without it
  Hint: once the session starts, ask goose to help debug the 'safeoutputs' extension
Warning: Failed to start extension 'github' (failed to initialize MCP client: Send message error Transport [rmcp::transport::worker::WorkerTransport<rmcp::transport::streamable_http_client::StreamableHttpClientWorker<reqwest::async_impl::client::Client>>] error: Client error: error sending request for url (http://localhost:8080/mcp/github), when send initialize request), continuing without it
  Hint: once the session starts, ask goose to help debug the 'github' extension

    __( O)>  ● new session · openai copilot/claude-sonnet-4.5
   \____)    20260804_1 · /home/runner/work/gh-aw/gh-aw
     L L     goose is ready
Ran into this error: Request failed: Resource not found (404) at http://172.30.0.30:10002/v1/chat/completions: .

Please retry if you think this is a transient or recoverable error.

[pelikhan]

@copilot the harness driver should fetch the /reflect endpoint and determine the available endpoints.

[Copilot]

🟡 Not ready to approve

The static endpoint regresses OpenAI/Codex routing in the shared engine.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

This review doesn't count toward merge requirements. Sign up for the private preview to control whether Copilot approvals count.

Pull request overview

Updates Goose CLI endpoint configuration to restore Copilot token telemetry.

Changes:

• Replaces ignored desktop-only variables with Goose CLI variables.
• Regenerates the Smoke Goose workflow.

File summaries

File	Description
.github/workflows/shared/goose.md	Configures the Goose OpenAI endpoint.
.github/workflows/smoke-goose.lock.yml	Regenerates the compiled workflow.

Review details

• Files reviewed: 2/2 changed files
• Comments generated: 1
• Review effort level: Balanced

We're testing this review assessment. Please use 👍 or 👎 to tell us if it's correct.

[gh-aw-bot]

@copilot Please address the latest review feedback on this PR, refresh the branch if GitHub allows it, and then run the pr-finisher skill before handing it back.

Open review feedback to close out (newest first):

• copilot-pull-request-reviewer: the static endpoint regresses OpenAI/Codex routing in the shared engine.

Please also recompile the workflow markdown after updating .github/workflows/shared/goose.md.

Generated by 👨🍳 PR Sous Chef

Generated by 👨‍🍳 PR Sous Chef · gpt54 · 8.54 AIC · ⌖ 6.21 AIC · ⊞ 8.3K · ◷
_{Comment /souschef to run again}

[pelikhan]

@copilot recompile and review this error:

Warning: Failed to start extension 'safeoutputs' (failed to initialize MCP client: Send message error Transport [rmcp::transport::worker::WorkerTransport<rmcp::transport::streamable_http_client::StreamableHttpClientWorkerreqwest::async_impl::client::Client>] error: Client error: error sending request for url (http://localhost:8080/mcp/safeoutputs), when send initialize request), continuing without it
Hint: once the session starts, ask goose to help debug the 'safeoutputs' extension
Warning: Failed to start extension 'github' (failed to initialize MCP client: Send message error Transport [rmcp::transport::worker::WorkerTransport<rmcp::transport::streamable_http_client::StreamableHttpClientWorkerreqwest::async_impl::client::Client>] error: Client error: error sending request for url (http://localhost:8080/mcp/github), when send initialize request), continuing without it
Hint: once the session starts, ask goose to help debug the 'github' extension

__( O)>  ● new session · openai copilot/claude-sonnet-4.5

____) 20260804_1 · /home/runner/work/gh-aw/gh-aw
L L goose is ready
Ran into this error: Request failed: Resource not found (404) at http://172.30.0.30:10002/v1/chat/completions: .

Harness should use /reflect to resolve the paths to the llms apis

[github-actions]

🎉 This pull request is included in a new release.

Release: v0.85.0