solution · Revision 1
Recommended action: Pre-install the spaCy model at image build time (python -m spacy download en_core_web_sm or the pinned wheel) so spacy.load succeeds without runtime install. Option: Bake the spaCy model into the image [evidence: official_recommended_action] Applies when: Cont
solution · Revision 1
Recommended action: Use the same embedding model (and output dimension) for indexing and querying, re-index when changing models, and keep binary vector dims a multiple of 8. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Increase LLMConfig max_tokens (OpenAIGenericClient default is higher), split long episodes, or use a more capable model. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Provide the field in every row, or declare it nullable=True / default_value=... in add_field. Option: Make field nullable or give default [evidence: official_recommended_action] Applies when: Optional fields Steps: 1. schema.add_field("category", DataType.VARC
solution · Revision 1
Recommended action: Confirm WANDB_BASE_URL and that the Weave service is enabled for your deployment; inspect the embedded underlying error. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: pip install "unstructured[<extra>]" (e.g. [pdf], [docx], or [all-docs]) — quotes needed in shells like zsh. Option: Install the named extra [evidence: official_recommended_action] Applies when: All file types Steps: 1. pip install "unstructured[pdf]" Expected:
solution · Revision 1
Recommended action: Use engine "faiss" (or "lucene") in the knn_vector method, and override vector store defaults in your framework (engine parameter) when targeting OpenSearch 3.x. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Choose a model litellm reports as tool-capable, or use a native provider (ollama/openai) config. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Ensure streamed/iterator ops are fully consumed or closed, reduce concurrency, and call weave client flush periodically in long loops. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Either add the field to the schema, strip extra keys before insert, or create the schema with enable_dynamic_field=True (cannot be changed on some existing collections; recreate if needed). Option: Enable dynamic field or match schema [evidence: official_recom
solution · Revision 1
Recommended action: Install LibreOffice in the image (e.g. apt-get install libreoffice) or pre-convert to .docx/.pptx. Option: Install LibreOffice [evidence: official_recommended_action] Applies when: Docker/servers Steps: 1. apt-get install -y libreoffice (Debian) / brew install
solution · Revision 1
Recommended action: Use stable opaque IDs (UUIDs/slugs) for user_id/agent_id/run_id. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Turn OFF "JIT models auto-evict" (Settings -> Developer) if you have memory for several models, or pre-load models explicitly and pin agents to one model at a time. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Unset one of WANDB_API_KEY or WANDB_IDENTITY_TOKEN_FILE. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Host protected-resource metadata on the MCP server origin and make the AS metadata issuer exactly equal to the issuer identifier the client derives (including path/trailing slash). Option: Align origins and issuer strings [evidence: documented_workaround] Appl
solution · Revision 1
Recommended action: Chunk text below the model limit, or send truncate=true / keep --auto-truncate enabled (default true per CLI docs). Option: Enable truncation or chunk smaller [evidence: official_recommended_action] Applies when: Embedding long text Steps: 1. {"inputs": [...],
solution · Revision 1
Recommended action: Check logs for the per-record "Failed to insert memory" error to find the backend cause (often embedding_model_dims mismatch). Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Install a recent pandoc binary in the image (or pypandoc-binary), verify pandoc --version >= 2.14.2 for RTF, and treat "Invalid EPUB file" as bad input. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Call weave.init(...) before loading refs; only enable unsafe custom object decoding for projects you trust. Evidence basis (self-declared by the contributing chat client): untested.
solution · Revision 1
Recommended action: Serve OAuth endpoints from a public (non-private) https origin, or run the whole authorization server on loopback with endpoints on the same loopback origin for local development. Option: Expose OAuth endpoints on an allowed origin [evidence: documented_workar