feat: simplify rag tool by simply calling gptme-rag via subprocess

[ErikBjare]

Gets rid of a lot of unnecessary complexity.

---

Important

Simplifies RAG tool by using gptme-rag CLI via subprocess, removing internal RAG management.

• Behavior:
  • Replaces internal RAG management with gptme-rag CLI calls in rag.py.
  • Removes _rag_context.py, eliminating internal RAG management classes and functions.
  • Updates rag_enhance_messages() to use gptme-rag CLI for context enhancement.
• Dependencies:
  • Removes gptme-rag from pyproject.toml dependencies.
• Tests:
  • Updates tests in test_tools_rag.py to mock subprocess.run for gptme-rag CLI calls.
  • Removes tests related to internal RAG management.

^{This description was created by for e929f30. It will automatically update as commits are pushed.}

[ellipsis-dev]

👍 Looks good to me! Reviewed everything up to cd31215 in 50 seconds

More details

• Looked at 588 lines of code in 6 files
• Skipped 0 files when reviewing.
• Skipped posting 2 drafted comments based on config settings.

1. gptme/tools/rag.py:76

• Draft comment:
  Consider handling FileNotFoundError in _run_rag_cmd to log an appropriate error message when gptme-rag is not found.
• Reason this comment was not posted:
  Confidence changes required: 50%
  The function _run_rag_cmd in gptme/tools/rag.py does not handle the case where the gptme-rag command is not found. This could lead to an unhandled FileNotFoundError. It would be better to catch this exception and log an appropriate error message.

2. gptme/tools/rag.py:161

• Draft comment:
  Handle the case where _run_rag_cmd(cmd) raises an exception in rag_enhance_messages to ensure the function returns a valid message even if the command fails.
• Reason this comment was not posted:
  Confidence changes required: 50%
  The rag_enhance_messages function in gptme/tools/rag.py uses _run_rag_cmd(cmd).stdout to get the command output, but if the command fails, it will raise an exception and not return any output. It would be better to handle this case and return an empty string or a default message instead.

Workflow ID: wflow_SkjNFJ2IZh0kF9tS

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[ellipsis-dev]

👍 Looks good to me! Incremental review on 89d29e7 in 17 seconds

More details

• Looked at 32 lines of code in 1 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. pyproject.toml:68

• Draft comment:
  The 'gptme-rag' dependency is removed from the extras but still present in the 'tool.poetry.dependencies' section. Consider removing it from here as well for consistency.
• Reason this comment was not posted:
  Comment did not seem useful.

Workflow ID: wflow_xzLx3uQDUZx8HDhM

---

You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 31.48148% with 37 lines in your changes missing coverage. Please review.

Project coverage is 72.37%. Comparing base (c0eb21f) to head (9b1b6f6).

✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
gptme/tools/rag.py	28.84%	37 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #316      +/-   ##
==========================================
- Coverage   73.56%   72.37%   -1.20%     
==========================================
  Files          68       67       -1     
  Lines        4975     4908      -67     
==========================================
- Hits         3660     3552     -108     
- Misses       1315     1356      +41     

Flag	Coverage Δ
anthropic/claude-3-haiku-20240307	70.55% <31.48%> (-1.11%)	⬇️
openai/gpt-4o-mini	69.58% <31.48%> (-0.90%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[ellipsis-dev]

❌ Changes requested. Incremental review on e929f30 in 35 seconds

More details

• Looked at 46 lines of code in 1 files
• Skipped 0 files when reviewing.
• Skipped posting 1 drafted comments based on config settings.

1. tests/test_tools_rag.py:25

• Draft comment:
  The docstring is inconsistent with the test logic. It should reflect that the test checks enhancement without RAG available, not with.
• Reason this comment was not posted:
  Comment was on unchanged code.

Workflow ID: wflow_Klu0CfrvB49cf7wR

---

Want Ellipsis to fix these issues? Tag @ellipsis-dev in a comment. You can customize Ellipsis with 👍 / 👎 feedback, review rules, user-specific overrides, quiet mode, and more.