What this error means
failed to parse introspection response: json: cannot unmarshal string into Go struct field .exp of type int64 is a Claude Code failure pattern reported for developers trying to fix claude code / claude.ai oauth authentication failure when using google as idp via mcp toolbox. Based on the imported evidence, treat this as a tool-specific troubleshooting page rather than a generic API error.
Why this happens
Issue #3240 in googleapis/mcp-toolbox (opened 2026-05-15): MCP auth with Google OIDC provider blocks all requests. Google oauth2/v3/tokeninfo returns numeric fields as quoted strings (exp, expires_in) but the Go struct declares Exp as int64 causing unmarshal panic → HTTP 500. Blocks end-to-end MCP OAuth flow documented in official docs. Category: AI Coding Tools because this directly affects Claude Code MCP connectivity — the primary user-facing symptom is Claude Code failing to authenticate.
Common causes
- Issue #3240 in googleapis/mcp-toolbox (opened 2026-05-15): MCP auth with Google OIDC provider blocks all requests. Google oauth2/v3/tokeninfo returns numeric fields as quoted strings (exp, expires_in) but the Go struct declares Exp as int64 causing unmarshal panic → HTTP 500. Blocks end-to-end MCP OAuth flow documented in official docs. Category: AI Coding Tools because this directly affects Claude Code MCP connectivity — the primary user-facing symptom is Claude Code failing to authenticate.
Quick fixes
- Confirm the exact error signature matches
failed to parse introspection response: json: cannot unmarshal string into Go struct field .exp of type int64. - Check the Claude Code account, local tool state, and provider configuration involved in the failing workflow.
- Verify the account session, API key, provider settings, and environment where the failing tool is running.
Platform/tool-specific checks
- Verify the command, editor, extension, or API client that produced the error.
- Compare local settings with CI, deployment, or editor-level settings when the error appears in only one environment.
- Avoid deleting credentials, local model data, or project settings until the failing scope is clear.
Step-by-step troubleshooting
- Capture the exact error message and the command, editor action, or request that triggered it.
- Check whether the failure is account/auth, quota/rate, model/provider, local runtime, or deployment configuration.
- Review the source evidence below and compare it with your environment.
- Apply one change at a time and rerun the smallest failing action.
- Keep the working fix documented for the team or deployment environment.
How to prevent it
- Keep provider/tool configuration documented.
- Record non-secret diagnostics such as tool version, provider name, model name, and command path.
- Add a lightweight check before CI or production workflows depend on the tool.