[aw] GitHub MCP Structural Analysis failed

[github-actions]

Workflow Failure

Workflow: GitHub MCP Structural Analysis
Branch: main
Run: https://github.com/github/gh-aw/actions/runs/24509507969

⚠️ Engine Failure: The claude engine terminated unexpectedly.

Last agent output:

2026-04-16T12:25:41.090Z [DEBUG] [API:request] Creating client, ANTHROPIC_CUSTOM_HEADERS present: false, has Authorization header: false
2026-04-16T12:25:41.090Z [DEBUG] [API:auth] OAuth token check starting
2026-04-16T12:25:41.117Z [DEBUG] [API:auth] OAuth token check complete
2026-04-16T12:25:41.118Z [DEBUG] Fast mode unavailable: Fast mode is not available in the Agent SDK
2026-04-16T12:25:41.144Z [DEBUG] [API REQUEST] /v1/messages source=sdk
2026-04-16T12:25:42.539Z [DEBUG] Stream started - received first chunk
{"type":"assistant","message":{"model":"claude-sonnet-4-6","id":"msg_015kiwtu461BmAB3uXrEZ2Mm","type":"message","role":"assistant","content":[{"type":"text","text":"Analysis complete. Here's what was done:\n\n**Phase 1 — Historical data**: Loaded 9 existing records from 2026-04-15.\n\n**Phase 2 — Tool testing**: Tested 9 representative tools across 9 toolsets. Key findings for today:\n- `list_discussions` (5/5) — best: 125 tokens, clean cursor pagination\n- `list_pull_requests` (4/5) — well structured, full body included\n- `list_issues` (4/5) — good; body inflated today by debug log content\n- `list_code_scanning_alerts` (2/5) — 145k chars (~36k tokens), unusable as-is\n- `search_code` (2/5) — repository sub-object embeds 30+ URL templates per result\n- `get_me` (1/5) — 403 blocked by integration token (same as yesterday)\n\n**Phase 3 — Data saved**: 9 new records appended to `/tmp/gh-aw/cache-memory/mcp_analysis.jsonl` (18 total, 2 days).\n\n**Phase 4 — Charts generated**: 4 charts (toolset sizes, usefulness ratings, daily trend, size vs usefulness).\n\n**Phase 5 — Discussion created** in the `audits` category with full analysis, tables, and embedded charts."}],"stop_reason":null,"stop_sequence":null,"stop_details":null,"usage":{"input_tokens":1,"cache_creation_input_tokens":202,"cache_read_input_tokens":87487,"cache_creation":{"ephemeral_5m_input_tokens":202,"ephemeral_1h_input_tokens":0},"output_tokens":1,"service_tier":"standard","inference_geo":"global"},"context_management":null},"parent_tool_use_id":null,"session_id":"2f670220-e303-4949-98ef-6627501f41d5","uuid":"a9858443-c881-4cca-873b-a2e47af5cf4c"}
2026-04-16T12:25:47.644Z [DEBUG] Fast mode unavailable: Fast mode is not available in the Agent SDK
{"type":"result","subtype":"success","is_error":false,"duration_ms":702488,"duration_api_ms":219226,"num_turns":37,"result":"Analysis complete. Here's what was done:\n\n**Phase 1 — Historical data**: Loaded 9 existing records from 2026-04-15.\n\n**Phase 2 — Tool testing**: Tested 9 representative tools across 9 toolsets. Key findings for today:\n- `list_discussions` (5/5) — best: 125 tokens, clean cursor pagination\n- `list_pull_requests` (4/5) — well structured, full body included\n- `list_issues` (4/5) — good; body inflated today by debug log content\n- `list_code_scanning_alerts` (2/5) — 145k chars (~36k tokens), unusable as-is\n- `search_code` (2/5) — repository sub-object embeds 30+ URL templates per result\n- `get_me` (1/5) — 403 blocked by integration token (same as yesterday)\n\n**Phase 3 — Data saved**: 9 new records appended to `/tmp/gh-aw/cache-memory/mcp_analysis.jsonl` (18 total, 2 days).\n\n**Phase 4 — Charts generated**: 4 charts (toolset sizes, usefulness ratings, daily trend, size vs usefulness).\n\n**Phase 5 — Discussion created** in the `audits` category with full analysis, tables, and embedded charts.","stop_reason":"end_turn","session_id":"2f670220-e303-4949-98ef-6627501f41d5","total_cost_usd":1.1565844500000002,"usage":{"input_tokens":67,"cache_creation_input_tokens":87689,"cache_read_input_tokens":2082349,"output_tokens":13523,"server_tool_use":{"web_search_requests":0,"web_fetch_requests":0},"service_tier":"standard","cache_creation":{"ephemeral_1h_input_tokens":0,"ephemeral_5m_input_tokens":87689},"inference_geo":"","iterations":[],"speed":"standard"},"modelUsage":{"claude-sonnet-4-6":{"inputTokens":67,"outputTokens":13523,"cacheReadInputTokens":2082349,"cacheCreationInputTokens":87689,"webSearchRequests":0,"costUSD":1.1565844500000002,"contextWindow":200000,"maxOutputTokens":32000}},"permission_denials":[],"terminal_reason":"completed","fast_mode_state":"off","uuid":"d3ae48c8-f134-4061-a08a-539c1bd5b86f"}
2026-04-16T12:29:04.966Z [DEBUG] Metrics export disabled by organization setting

Action Required

Assign this issue to Copilot using the agentic-workflows sub-agent to automatically debug and fix the workflow failure.

Debug with any coding agent

Use this prompt with any coding agent (GitHub Copilot, Claude, Gemini, etc.):

Debug the agentic workflow failure using https://raw.githubusercontent.com/github/gh-aw/main/debug.md

The failed workflow run is at https://github.com/github/gh-aw/actions/runs/24509507969

Manually invoke the agent

Debug this workflow failure using your favorite Agent CLI and the agentic-workflows prompt.

• Start your agent
• Load the agentic-workflows prompt from .github/agents/agentic-workflows.agent.md or https://github.com/github/gh-aw/blob/main/.github/agents/agentic-workflows.agent.md
• Type debug the agentic workflow github-mcp-structural-analysis failure in https://github.com/github/gh-aw/actions/runs/24509507969

Tip

Stop reporting this workflow as a failure

To stop a workflow from creating failure issues, set report-failure-as-issue: false in its frontmatter:

safe-outputs:
  report-failure-as-issue: false

Generated from GitHub MCP Structural Analysis · ◷

• expires on Apr 23, 2026, 12:31 PM UTC

[pelikhan]

Create a PR that adds a JavaScript agent driver for Claude. See how the copilot driver is done.

• refactor copilot driver to keep logic
• configured regexes to detect various error, but keep same retry logic
• refactor how to launch agent