[BUG] `TypeError: TextInputSequence must be str` when used with `codecompanion.nvim`

[Bekaboo]

Describe the bug

When codecompanion.nvim uses the provided @vectorcode tools as described in the doc, I get the following error:

Traceback (most recent call last):
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/main.py", line 78, in async_main
    return_val = await query(final_configs)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py", line 185, in query
    structured_result = await build_query_results(collection, configs)
                        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py", line 82, in build_query_results
    for identifier in await get_query_result_files(collection, configs):
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py", line 74, in get_query_result_files
    ).rerank(results)
      ~~~~~~^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/reranker.py", line 73, in rerank
    ranks = self.model.rank(
        self.query_chunks[query_chunk_idx], chunk_docs, apply_softmax=True
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transformers/cross_encoder/util.py", line 68, in wrapper
    return func(self, *args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transformers/cross_encoder/CrossEncoder.py", line 500, in rank
    scores = self.predict(
        sentences=query_doc_pairs,
    ...<5 lines>...
        convert_to_tensor=convert_to_tensor,
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transformers/cross_encoder/util.py", line 68, in wrapper
    return func(self, *args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transformers/cross_encoder/CrossEncoder.py", line 399, in predict
    features = self.tokenizer(
        batch,
    ...<2 lines>...
        return_tensors="pt",
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_base.py", line 2887, in __call__
    encodings = self._call_one(text=text, text_pair=text_pair, **all_kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_base.py", line 2975, in _call_one
    return self.batch_encode_plus(
           ~~~~~~~~~~~~~~~~~~~~~~^
        batch_text_or_text_pairs=batch_text_or_text_pairs,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<17 lines>...
        **kwargs,
        ^^^^^^^^^
    )
    ^
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_base.py", line 3177, in batch_encode_plus
    return self._batch_encode_plus(
           ~~~~~~~~~~~~~~~~~~~~~~~^
        batch_text_or_text_pairs=batch_text_or_text_pairs,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<17 lines>...
        **kwargs,
        ^^^^^^^^^
    )
    ^
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_fast.py", line 539, in _batch_encode_plus
    encodings = self._tokenizer.encode_batch(
        batch_text_or_text_pairs,
        add_special_tokens=add_special_tokens,
        is_pretokenized=is_split_into_words,
    )
TypeError: TextInputSequence must be str

To Reproduce
Steps to reproduce the behavior:

1. :CodeCompanionChat to open chat buffer in neovim
2. Goto some vectorised repo (in this case the repo of VectorCode itself), start the chat with @full_stack_dev @vectorcode Briefly summarize the project for me
3. CodeCompanion uses the vectorcode tool with the following XML:

   <tools>
     <tool name="vectorcode">
       <action>
         <command>query</command>
         <options>
           <query>README</query>
           <query>documentation</query>
           <project_root>/home/user/.local/share/nvim/packages/VectorCode</project_root>
           <count>10</count>
         </options>
       </action>
     </tool>
   </tools>

4. Error

Expected behavior
Should not error.

VectorCode Configuration
Please attach your <project_root>/.vectorcode/config.json or
~/.config/vectorcode/config.json here.

No such file.

For issues with the Neovim plugin, please also attach your setup options:

No setup() called.

If it only occurs when you use VectorCode with a particular plugin, please
attach the relevant config here:

-- VectorCode integration to provide project-level context
local vc = vim.fn.executable('vectorcode') == 1
  and vim.F.npcall(require, 'vectorcode.integrations')

require('codecompanion').setup({
  strategies = {
    chat = {
      adapter = adapter,
      slash_commands = {
        ['codebase'] = vc and vc.codecompanion.chat.make_slash_command(),
      },
      tools = {
        vectorcode = vc
          and {
            description = 'Run VectorCode to retrieve the project context.',
            callback = vc.codecompanion.chat.make_tool(),
          },
      },
  },
})

Platform information:

• If the issue is about the CLI, attach the output of pipx runpip vectorcode freeze:

annotated-types==0.7.0
anyio==4.9.0
asgiref==3.8.1
backoff==2.2.1
bcrypt==4.3.0
build==1.2.2.post1
cachetools==5.5.2
certifi==2025.1.31
charset-normalizer==3.4.1
chroma-hnswlib==0.7.6
chromadb==0.6.3
click==8.1.8
coloredlogs==15.0.1
Deprecated==1.2.18
distro==1.9.0
durationpy==0.9
fastapi==0.115.12
filelock==3.18.0
flatbuffers==25.2.10
fsspec==2025.3.2
google-auth==2.38.0
googleapis-common-protos==1.69.2
grpcio==1.71.0
h11==0.14.0
httpcore==1.0.7
httptools==0.6.4
httpx==0.28.1
huggingface-hub==0.30.1
humanfriendly==10.0
idna==3.10
importlib_metadata==8.6.1
importlib_resources==6.5.2
Jinja2==3.1.6
joblib==1.4.2
kubernetes==32.0.1
markdown-it-py==3.0.0
MarkupSafe==3.0.2
mdurl==0.1.2
mmh3==5.1.0
monotonic==1.6
mpmath==1.3.0
networkx==3.4.2
numpy==2.2.4
nvidia-cublas-cu12==12.4.5.8
nvidia-cuda-cupti-cu12==12.4.127
nvidia-cuda-nvrtc-cu12==12.4.127
nvidia-cuda-runtime-cu12==12.4.127
nvidia-cudnn-cu12==9.1.0.70
nvidia-cufft-cu12==11.2.1.3
nvidia-curand-cu12==10.3.5.147
nvidia-cusolver-cu12==11.6.1.9
nvidia-cusparse-cu12==12.3.1.170
nvidia-cusparselt-cu12==0.6.2
nvidia-nccl-cu12==2.21.5
nvidia-nvjitlink-cu12==12.4.127
nvidia-nvtx-cu12==12.4.127
oauthlib==3.2.2
onnxruntime==1.21.0
opentelemetry-api==1.31.1
opentelemetry-exporter-otlp-proto-common==1.31.1
opentelemetry-exporter-otlp-proto-grpc==1.31.1
opentelemetry-instrumentation==0.52b1
opentelemetry-instrumentation-asgi==0.52b1
opentelemetry-instrumentation-fastapi==0.52b1
opentelemetry-proto==1.31.1
opentelemetry-sdk==1.31.1
opentelemetry-semantic-conventions==0.52b1
opentelemetry-util-http==0.52b1
orjson==3.10.16
overrides==7.7.0
packaging==24.2
pathspec==0.12.1
pillow==11.1.0
posthog==3.23.0
protobuf==5.29.4
psutil==7.0.0
pyasn1==0.6.1
pyasn1_modules==0.4.2
pydantic==2.11.1
pydantic_core==2.33.0
Pygments==2.19.1
PyPika==0.48.9
pyproject_hooks==1.2.0
python-dateutil==2.9.0.post0
python-dotenv==1.1.0
PyYAML==6.0.2
regex==2024.11.6
requests==2.32.3
requests-oauthlib==2.0.0
rich==14.0.0
rsa==4.9
safetensors==0.5.3
scikit-learn==1.6.1
scipy==1.15.2
sentence-transformers==4.0.1
setuptools==78.1.0
shellingham==1.5.4
shtab==1.7.1
six==1.17.0
sniffio==1.3.1
starlette==0.46.1
sympy==1.13.1
tabulate==0.9.0
tenacity==9.0.0
threadpoolctl==3.6.0
tokenizers==0.21.1
torch==2.6.0
tqdm==4.67.1
transformers==4.50.3
tree-sitter==0.24.0
tree-sitter-c-sharp==0.23.1
tree-sitter-embedded-template==0.23.2
tree-sitter-language-pack==0.6.1
tree-sitter-yaml==0.7.0
triton==3.2.0
typer==0.15.2
typing-inspection==0.4.0
typing_extensions==4.13.0
urllib3==2.3.0
uvicorn==0.34.0
uvloop==0.21.0
-e git+https://github.com/Davidyz/VectorCode.git@4775829c86971f6e2a77f556619a004e324ee4d9#egg=VectorCode
watchfiles==1.0.4
websocket-client==1.8.0
websockets==15.0.1
wrapt==1.17.2
zipp==3.21.0

• If the issue is about the neovim plugin, attach the neovim version you're using: 0.11

System Information:

• OS: Linux

Additional context
Add any other context about the problem here.

[Davidyz]

Hi, thanks for reporting this error. Does this only occur when using with the full_stack_dev tool?

[Bekaboo]

No. It also happens without the @full_stack_dev tool. Here is the full conversation:

## Me

> Sharing:
> - <tool>vectorcode</tool>

@vectorcode briefly summarize vectorcode project for me

## CodeCompanion (Anthropic)

### Reasoning

The user wants me to use the VectorCode tool to retrieve information about the VectorCode project itself and provide a summary. I'll need to:

1. First run the `ls` command to get a list of indexed projects
2. Then query for relevant files in the vectorcode project
3. Analyze the results and provide a concise summary

Let me start by running the `ls` command:

### Response

\`\`\`xml
<tools>
  <tool name="vectorcode">
    <action>
      <command>ls</command>
    </action>
  </tool>
</tools>
\`\`\`

I'll retrieve information about the VectorCode project. Please wait for the results to be fetched.

## Me

> Sharing:
> - <tool>vectorcode</tool>



## CodeCompanion (Anthropic)

### Reasoning

Based on the results returned by the ls command, I can see that there are three indexed projects:
1. /home/user/xxxxx
2. /home/user/xxxxx
3. /home/user/.local/share/nvim/packages/VectorCode

From these, the most relevant one for understanding the VectorCode project itself would be the last one: `/home/user/.local/share/nvim/packages/VectorCode`. This appears to be the actual location of the VectorCode tool that we're trying to learn about.

Let me use the VectorCode tool to query information about itself from its own repository. I'll start by retrieving some general information:

### Response

Now I'll query the VectorCode project to get a summary of what it is:

\`\`\`xml
<tools>
  <tool name="vectorcode">
    <action>
      <options>
        <query>README</query>
        <query>description</query>
        <query>about</query>
        <project_root>/home/user/.local/share/nvim/packages/VectorCode</project_root>
        <count>10</count>
      </options>
      <command>query</command>
    </action>
  </tool>
</tools>
\`\`\`

## Me

> Sharing:
> - <tool>vectorcode</tool>

Then I get error:

Traceback (most recent call last):
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/main.py", line 78, in async_main
    return_val = await query(final_configs)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py", line 185, in query
    structured_result = await build_query_results(collection, configs)
                        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py", line 82, in build_query_results
    for identifier in await get_query_result_files(collection, configs):
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py", line 74, in get_query_result_files
    ).rerank(results)
      ~~~~~~^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/reranker.py", line 73, in rerank
    ranks = self.model.rank(
        self.query_chunks[query_chunk_idx], chunk_docs, apply_softmax=True
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transformers/cross_encoder/util.py", line 68, in wrapper
    return func(self, *args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transformers/cross_encoder/CrossEncoder.py", line 500, in rank
    scores = self.predict(
        sentences=query_doc_pairs,
    ...<5 lines>...
        convert_to_tensor=convert_to_tensor,
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transformers/cross_encoder/util.py", line 68, in wrapper
    return func(self, *args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transformers/cross_encoder/CrossEncoder.py", line 399, in predict
    features = self.tokenizer(
        batch,
    ...<2 lines>...
        return_tensors="pt",
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_base.py", line 2887, in __call__
    encodings = self._call_one(text=text, text_pair=text_pair, **all_kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_base.py", line 2975, in _call_one
    return self.batch_encode_plus(
           ~~~~~~~~~~~~~~~~~~~~~~^
        batch_text_or_text_pairs=batch_text_or_text_pairs,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<17 lines>...
        **kwargs,
        ^^^^^^^^^
    )
    ^
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_base.py", line 3177, in batch_encode_plus
    return self._batch_encode_plus(
           ~~~~~~~~~~~~~~~~~~~~~~~^
        batch_text_or_text_pairs=batch_text_or_text_pairs,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<17 lines>...
        **kwargs,
        ^^^^^^^^^
    )
    ^
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_fast.py", line 539, in _batch_encode_plus
    encodings = self._tokenizer.encode_batch(
        batch_text_or_text_pairs,
        add_special_tokens=add_special_tokens,
        is_pretokenized=is_split_into_words,
    )
TypeError: TextInputSequence must be str

[Davidyz]

I wasn't able to reproduce this error, but I added explicit conversion for the query results to make sure they're all str when passed to the reranker model in this PR. Could you give it a try and see if it fix the problem?

If it still doesn't work in codecompanion, could you try making queries from the CLI (in the project directory) and see if it throws the same error?

[Bekaboo]

The error still occurs.

If it still doesn't work in codecompanion, could you try making queries from the CLI (in the project directory) and see if it throws the same error?

I tried with CLI and get the same error:

$ vectorcode query jobs

Host at 127.0.0.1:8000 is unavailable. VectorCode will start its own Chromadb at a random port.
Starting querying...
Traceback (most recent call last):
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/main.py", line 78, in async_ma
in
    return_val = await query(final_configs)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py"
, line 185, in query
    structured_result = await build_query_results(collection, configs)
                        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py"
, line 82, in build_query_results
    for identifier in await get_query_result_files(collection, configs):
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/__init__.py"
, line 74, in get_query_result_files
    ).rerank(results)
      ~~~~~~^^^^^^^^^
  File "/home/user/.local/share/nvim/packages/VectorCode/src/vectorcode/subcommands/query/reranker.py"
, line 73, in rerank
    ranks = self.model.rank(
        self.query_chunks[query_chunk_idx], chunk_docs, apply_softmax=True
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transforme
rs/cross_encoder/util.py", line 68, in wrapper
    return func(self, *args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transforme
rs/cross_encoder/CrossEncoder.py", line 500, in rank
    scores = self.predict(
        sentences=query_doc_pairs,
    ...<5 lines>...
        convert_to_tensor=convert_to_tensor,
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/torch/utils/_contex
tlib.py", line 116, in decorate_context
    return func(*args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transforme
rs/cross_encoder/util.py", line 68, in wrapper
    return func(self, *args, **kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/sentence_transforme
rs/cross_encoder/CrossEncoder.py", line 399, in predict
    features = self.tokenizer(
        batch,
    ...<2 lines>...
        return_tensors="pt",
    )
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_b
ase.py", line 2887, in __call__
    encodings = self._call_one(text=text, text_pair=text_pair, **all_kwargs)
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_b
ase.py", line 2975, in _call_one
    return self.batch_encode_plus(
           ~~~~~~~~~~~~~~~~~~~~~~^
        batch_text_or_text_pairs=batch_text_or_text_pairs,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<17 lines>...
        **kwargs,
        ^^^^^^^^^
    )
    ^
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_b
ase.py", line 3177, in batch_encode_plus
    return self._batch_encode_plus(
           ~~~~~~~~~~~~~~~~~~~~~~~^
        batch_text_or_text_pairs=batch_text_or_text_pairs,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    ...<17 lines>...
        **kwargs,
        ^^^^^^^^^
    )
    ^
  File "/home/user/.local/share/pipx/venvs/vectorcode/lib/python3.13/site-packages/transformers/tokenization_utils_f
ast.py", line 539, in _batch_encode_plus
    encodings = self._tokenizer.encode_batch(
        batch_text_or_text_pairs,
        add_special_tokens=add_special_tokens,
        is_pretokenized=is_split_into_words,
    )
TypeError: TextInputSequence must be str

[Davidyz]

I still couldn't reproduce this, and I found this issue over in transformers repo (this is used under-the-hood by the default embedding function/reranker model): huggingface/transformers#28116, which is still not fully resolved yet, so I don't think this issue can be resolved any time soon.

In the mean time, you can set reranker to null (if you've never created a project-local config, you can do this in the global configuration at ~/.config/vectorcode/config.json). This will disable the sentence transformers reranker and should bypass this error.

{
  "reranker": null
}