Summary
This issue is documentation-oriented.
In real usage, when MCP execution has a startup/runtime problem, agents may continue by using direct txc CLI commands in terminal. That fallback behavior is already happening in practice.
Request
Please document this behavior so contributors and users understand what to expect:
- When MCP operations fail to execute, agent workflows may use direct
txc CLI as a practical fallback.
- This is expected operational behavior for continuity, not an unexpected workaround.
- Output/trace implications should be described (MCP result vs terminal command output).
- Include guidance for Windows/macOS/Linux environments.
Why this matters
- Reduces confusion during troubleshooting.
- Helps maintainers and contributors interpret agent actions correctly.
- Makes the supported operational model explicit.
Suggested documentation locations
README.md
- MCP docs / troubleshooting docs (for example in
docs/)
- Any agent/MCP usage guidance
Environment context
- Observed in Windows workflows
- Applies to cross-platform usage (Windows/macOS/Linux)
- Repo: TALXIS/tools-cli
Summary
This issue is documentation-oriented.
In real usage, when MCP execution has a startup/runtime problem, agents may continue by using direct
txcCLI commands in terminal. That fallback behavior is already happening in practice.Request
Please document this behavior so contributors and users understand what to expect:
txcCLI as a practical fallback.Why this matters
Suggested documentation locations
README.mddocs/)Environment context