● Community
cocoindex-io
⚡ Instant
cocoindex-claude
by cocoindex-io · cocoindex-io/cocoindex-claude
Teaches Claude Code how to build CocoIndex data-transformation pipelines — real-time incremental processing for vector DBs and knowledge graphs.
CocoIndex is a real-time data transformation framework for AI pipelines (source → transform → vector DB / knowledge graph). This skill loads its conventions into Claude Code so the model writes idiomatic CocoIndex flows, custom functions, and CLI/API commands instead of guessing at an unfamiliar library.
Why use it
Key features
- Drop-in knowledge of CocoIndex flow definitions, sources, and sinks
- Guidance on writing custom transformation functions correctly
- Claude picks the right CLI vs programmatic API based on task
- Understands incremental processing semantics so it doesn't suggest full rebuilds
- Official — authored by the CocoIndex team
Live Demo
What it looks like in practice
cocoindex-claude-skill.replay
▶ ready
Use Cases
Real-world ways to use cocoindex-claude
When to use: You've read the CocoIndex homepage and want Claude to scaffold a working flow instead of copy-pasting tutorials.
Flow
-
Describe the pipeline
Use the cocoindex skill — scaffold a flow that watches ./docs, chunks markdown, embeds with OpenAI, writes to Qdrant.✓ Copied
→ Claude generates a flow.py with @cocoindex.flow_def and the right source/transform/export calls
-
Run and verify
Now show me the exact CLI commands to build, run, and incrementally update this flow.✓ Copied
→ cocoindex setup / update / eval commands in the correct order
Outcome: A working flow.py plus the commands to run it — no more guessing at the CocoIndex decorator surface.
Pitfalls
- Claude invents APIs that don't exist in CocoIndex — Explicitly prompt 'use only APIs documented in the cocoindex skill' — hallucination drops when the skill is in scope
When to use: Built-in transforms don't cover your parsing / enrichment step.
Flow
-
Describe the transform
With the cocoindex skill — write a @cocoindex.op.function that takes a PDF path, extracts tables, returns structured rows.✓ Copied
→ Idiomatic function signature with correct typed inputs/outputs
-
Wire it into the flow
Now plug this into my existing flow at the chunking stage.✓ Copied
→ Diff against current flow.py with the new op in place
Outcome: A custom transform that obeys CocoIndex's incremental semantics.
Pitfalls
- Function returns non-deterministic output, breaking incremental processing — Skill reminds Claude that transforms must be pure — if you hit this, re-read the incremental section
Combinations
Pair with other MCPs for X10 leverage
Let Claude actually write the flow.py file to disk, not just print it in chat
Scaffold the CocoIndex flow and save it to ./pipelines/docs_to_qdrant.py.✓ Copied
Pipe CocoIndex output into a Postgres vector store and query it from the same session
Build the flow targeting pgvector, then run a retrieval query against it.✓ Copied
Cost & Limits
What this costs to run
- API quota
- None
- Tokens per call
- Modest — skill loads reference material only when relevant
- Monetary
- Free — skills are local files
- Tip
- CocoIndex itself is free; costs come from your embedding model and vector DB, not this skill.
Security
Permissions, secrets, blast radius
Credential storage: No credentials — pure prompts
Data egress: None from the skill itself. Your pipeline calls whichever embedding / DB services you configure.
Troubleshooting
Common errors and fixes
Wrong trigger keywords — Claude writes generic Python instead of CocoIndex
Mention CocoIndex explicitly, or say 'scaffold a cocoindex flow_def'.
Alternatives
cocoindex-claude vs others
| Alternative | When to use it instead | Tradeoff |
|---|
| LlamaIndex skills | You want LlamaIndex's retrieval abstractions instead of CocoIndex's transform-centric model | Different mental model — LlamaIndex is retrieval-first, CocoIndex is pipeline-first |
| Hand-written Python | You need one-off processing and don't want incremental state | No incremental rebuilds, no framework conventions |