fix: make built-in tool injection configurable

[yashhzd]

Summary

Fixes #90

Every ToolManager instance automatically copies the global tool registry (_GLOBAL_TOOL_REGISTRY), so all agents receive move_one_step, teleport_to_location, and speak_to regardless of whether those capabilities make sense for the agent. This causes:

• Capability leakage — static entities gain movement tools
• Tool noise — LLMs waste reasoning on irrelevant tools
• Hard-to-debug conflicts — custom tools can shadow built-ins
• No local permission reasoning — cannot restrict agent capabilities

Changes

tool_manager.py:

• Add include_builtins parameter to ToolManager.__init__() (default True for full backward compatibility)
• When False, the manager starts with an empty tool set — only extra_tools (if provided) are registered
• Add remove_tool(name) method for fine-grained per-instance tool removal

llm_agent.py:

• Add include_builtin_tools parameter to LLMAgent.__init__() that passes through to ToolManager

Tests (5 new):

• test_include_builtins_false_starts_empty — verifies no global tools when disabled
• test_include_builtins_true_includes_global_tools — verifies default behavior preserved
• test_include_builtins_false_with_extra_tools — verifies only extras are available
• test_remove_tool — verifies removal works
• test_remove_tool_missing_is_silent — verifies no error on missing tool

Usage Example

# Agent WITHOUT spatial tools (e.g., dining philosopher)
class Philosopher(LLMAgent):
    def __init__(self, model, **kwargs):
        super().__init__(model, include_builtin_tools=False, **kwargs)
        # Only register tools relevant to this agent
        self.tool_manager.register(pick_up_fork)
        self.tool_manager.register(put_down_fork)

# Agent WITH default tools (backward compatible, no change needed)
class Explorer(LLMAgent):
    def __init__(self, model, **kwargs):
        super().__init__(model, **kwargs)  # gets all built-ins

Test Plan

• All 267 existing tests pass (5 new)
• Pre-commit hooks pass (ruff, codespell)
• Fully backward compatible (default include_builtins=True)

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 410232f1-eb94-4339-864f-59c5b28ca122

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

Tip

Try Coding Plans. Let us write the prompt for your AI agent so you can ship faster (with fewer bugs).
Share your feedback on Discord.

---

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

• X
• Mastodon
• Reddit
• LinkedIn

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

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.09%. Comparing base (4c0549e) to head (56b4af3).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #167      +/-   ##
==========================================
+ Coverage   90.08%   90.09%   +0.01%     
==========================================
  Files          19       19              
  Lines        1503     1505       +2     
==========================================
+ Hits         1354     1356       +2     
  Misses        149      149              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.