Skip to content

Conversation

@schuellerf
Copy link
Contributor

@schuellerf schuellerf commented Nov 12, 2025

Update claude skill and related docs

I tried to align README, code and claude skill
Please check thoroughly if I understood the config correctly

Summary by CodeRabbit

  • Documentation

    • Updated MCP agent configuration documentation with new fields and command-line flags.
    • Updated evaluation command guidance and file format references.
  • Configuration

    • Added useVirtualHome and allowedToolsJoinSeparator configuration options for agents.
    • Added --strict-mcp-config and --allowedTools command-line flags.

@coderabbitai
Copy link

coderabbitai bot commented Nov 12, 2025

Walkthrough

Documentation updates across three files: skill documentation for eval file handling and gevals command, MCP agent configuration example in README with new fields (useVirtualHome, allowedToolsJoinSeparator) and flags (--strict-mcp-config, --allowedTools), and a config struct comment reflecting the template variable naming update.

Changes

Cohort / File(s) Summary
Skill and eval documentation
.claude/skills/create-eval/SKILL.md
Updated eval file reference from eval.yaml to eval.md in from-scratch creation flow; changed gevals command invocation from gevals run to gevals eval subcommand.
Configuration documentation
README.md, pkg/agent/config.go
Added new MCP agent configuration fields in README example: useVirtualHome: false and allowedToolsJoinSeparator: ",". Updated runPrompt command with flags --strict-mcp-config and --allowedTools "{{ .AllowedToolArgs }}". Updated documentation comment in AgentCommands struct to reflect template variable name change from {{ .AllowedTools }} to {{ .AllowedToolArgs }}.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

  • Verify consistency between README example and config.go comment template variable naming
  • Confirm eval file reference change (eval.yaml → eval.md) and gevals subcommand rename (run → eval) are correct and documented consistently
  • Check that new MCP agent configuration fields align with actual implementation expectations

Pre-merge checks and finishing touches

✅ Passed checks (3 passed)
Check name Status Explanation
Description Check ✅ Passed Check skipped - CodeRabbit’s high-level summary is enabled.
Title check ✅ Passed The title 'docs: update claude skill and related docs' accurately reflects the main changes: documentation updates to the claude skill in SKILL.md, README.md, and a comment in pkg/agent/config.go.
Docstring Coverage ✅ Passed No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
✨ Finishing touches
  • 📝 Generate docstrings
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Post copyable unit tests in a comment

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands and usage tips.

Copy link

@coderabbitai coderabbitai bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Actionable comments posted: 1

📜 Review details

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 912c90c and cf8e6e1.

📒 Files selected for processing (3)
  • .claude/skills/create-eval/SKILL.md (2 hunks)
  • README.md (1 hunks)
  • pkg/agent/config.go (1 hunks)
🔇 Additional comments (3)
.claude/skills/create-eval/SKILL.md (1)

30-30: LGTM! Command rename is consistent.

The command has been correctly updated from gevals run to gevals eval, which aligns with the usage shown in README.md line 24.

pkg/agent/config.go (1)

48-48: LGTM! Documentation aligned with README.

The comment has been correctly updated to reference {{ .AllowedToolArgs }}, which matches the template variable used in the README.md example (line 77). This improves consistency across the documentation.

README.md (1)

72-77: LGTM! Agent configuration example updated correctly.

The agent.yaml example has been enhanced with:

  • useVirtualHome: false - matches the struct field in config.go line 30
  • allowedToolsJoinSeparator: "," - matches the struct field in config.go line 43
  • Updated runPrompt with new flags --strict-mcp-config and --allowedTools "{{ .AllowedToolArgs }}" - aligns with the comment in config.go line 48

These changes properly document the new configuration options and demonstrate their usage.

Copy link
Contributor

@Cali0707 Cali0707 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for updating this @schuellerf !


```bash
gevals run <path to eval yaml file>
gevals eval <path to eval yaml file>
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nice catch, forgot to update this when we added #19

On that note, would you mind adding a line here explaining the -run flag? I'm assuming a lot of the time, claude will want to run only a subset of an eval suite

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants