fix: ReActReasoning.__repr__ AttributeError and STLTMemory Mesa 4.x model.steps compatibility

[abhinavk0220]

Fixes #205

Fixes #204

Summary

Two small but real bugs fixed in this PR:

1. ReActReasoning.__repr__() crashes with AttributeError
2. STLTMemory uses model.steps which does not exist in Mesa 4.x

---

Bug 1 — ReActReasoning.__repr__() AttributeError

ReActReasoning.__repr__() was referencing self.remaining_tool_calls
which does not exist on ReActReasoning. That attribute belongs to
ReWOOReasoning only. Calling repr() on any ReActReasoning instance
would crash with:

AttributeError: 'ReActReasoning' object has no attribute 'remaining_tool_calls'

Fix: Removed remaining_tool_calls from ReActReasoning.__repr__().
ReWOOReasoning.__repr__() correctly keeps it since that class does have
the attribute.

---

Bug 2 — STLTMemory uses model.steps (Mesa 4.x incompatibility)

st_lt_memory.py was using self.agent.model.steps to record the current
step when consolidating memory entries. model.steps does not exist in
Mesa 4.x — it was removed. This causes AttributeError during any
simulation run that triggers memory consolidation.

# Before — crashes in Mesa 4.x
step=self.agent.model.steps,

# After — correct Mesa 4.x API
step=int(getattr(self.agent.model, "_time", 0)),

The same fix was already applied to record_model.py — this brings
st_lt_memory.py in line with the same pattern.

---

Testing

All existing tests pass:

python -m pytest tests/ -q --ignore=tests/test_parallel_stepping.py
# 0 failures

---

GSoC contributor checklist

Context & motivation

Found these two bugs while auditing the codebase after fixing the Mesa 4.x
issues in PR #195. ReActReasoning.__repr__() was referencing
self.remaining_tool_calls which only exists on ReWOOReasoning calling
repr() on any ReActReasoning instance would crash silently. The
STLTMemory bug surfaced the same pattern as record_model.py model.steps
removed in Mesa 4.x but still used in the memory consolidation path.

What I learned

Small bugs in base classes and memory modules can be harder to catch than
bugs in tools because they only surface under specific conditions
__repr__ only when you inspect an agent, memory consolidation only when
short-term capacity is exceeded. Reading the codebase with fresh eyes after
fixing one bug is often how you find the next one.

Learning repo

🔗 My learning repo: https://github.com/abhinavk0220/GSoC-learning-space
🔗 Relevant model(s): https://github.com/abhinavk0220/GSoC-learning-space/tree/main/models

Readiness checks

• This PR addresses an agreed-upon problem (linked issue or discussion with maintainer approval), or is a small/trivial fix
• I have read the contributing guide and deprecation policy
• I have performed a self-review: I reviewed my own PR as if I were a reviewer and left comments on anything that needs explanation
• Another GSoC contributor has reviewed this PR:
• Tests pass locally (pytest --cov=mesa tests/)
• Code is formatted (ruff check . --fix)
• If applicable: documentation, examples, and/or migration guide are updated

[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: 8705859e-9344-4339-876b-4d926d7cd4ee

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

---

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.}

Tip

You can disable sequence diagrams in the walkthrough.

Disable the reviews.sequence_diagrams setting to disable sequence diagrams in the walkthrough.

[abhinavk0220]

Same pattern as record_model.py model.steps doesn't exist in Mesa 4.x. Using int(getattr(self.agent.model, '_time', 0)) as the safe fallback. This only surfaces at runtime when memory consolidation triggers, which is why it wasn't caught earlier.