Skip to main content

ty-find

An LSP adapter for AI coding agents. Symbol name in, structured code intelligence out.

LSP servers are the gold standard for code navigation — but they require file positions (file.py:29:7). LLMs think in symbol names (MyClass). To use an LSP, an LLM first has to grep for the position, which is imprecise and adds a round-trip. tyf bridges this gap: one command gives you definition, signature, and references — by name, no file paths needed.

$ tyf show list_animals
# Definition (func)
main.py:14:1

# Signature
def list_animals(animals: list[Animal]) -> None

# Refs: 2 across 1 file(s)

$ tyf show list_animals --all    # add docs, refs, test refs

Built for: Claude Code, Codex, Cursor, Gemini CLI — and humans who want fast terminal-based navigation.

Why tyf?

vs grep/ripgrep:

  • grep matches text — tyf understands Python's type system
  • grep returns hits in comments, strings, and docstrings; tyf returns only real symbol references

vs raw LSP (in editors):

  • LSP requires file:line:col positions to answer queries
  • An LLM doesn't know positions without searching first
  • Searching with grep is imprecise — circular problem
  • tyf accepts symbol names directly, resolves positions internally

Usage with Claude Code

Add this to your project's CLAUDE.md to enable type-aware code navigation:

### Python Symbol Navigation — `tyf`

This project has `tyf` — a type-aware code search that gives LSP-quality
results by symbol name. Use `tyf` instead of grep/ripgrep for Python symbol lookups.

- `tyf show my_function` — definition + signature (add `-d` docs, `-r` refs, `-t` test refs, or `--all`)
- `tyf find MyClass` — find definition location
- `tyf refs my_function` — all usages (before refactoring)
- `tyf members TheirClass` — class public API
- `tyf calls my_function` — call tree (`--in` for callers, before refactoring)
- `tyf list file.py` — file outline

All commands accept multiple symbols — batch to save tool calls.
Run `tyf <cmd> --help` for options.

Use grep for: string literals, config values, TODOs, non-Python files.

Installation

Prerequisite: ty type checker (uv add --dev ty)

Optional: ripgrep (rg) — speeds up lookups for non-existent symbols by quickly verifying whether a symbol appears in any .py file before retrying LSP queries. Without it, searches for non-existent symbols still work but may be slower.

# macOS
brew install ripgrep

# Ubuntu/Debian
sudo apt install ripgrep

# Or via cargo
cargo install ripgrep
uv add --dev ty-find

Not sure where to start? uvx ty-find guide prints agent-facing instructions (setup, then the command reference) without installing anything or starting the daemon. Inside a project that has it installed, uv run tyf guide does the same.

Note: On Windows, only tyf find --file and tyf guide are supported for now. All other commands require Unix domain sockets (Linux, macOS).

Usage

Show (Definition + Signature + References)

All-in-one command — searches the workspace by symbol name, no file needed. Add -d (docs), -r (references), -t (test refs), or --all for everything:

tyf show calculate_sum

# Multiple symbols at once
tyf show calculate_sum UserService Config

# Include docstring + refs + test refs
tyf show calculate_sum --all

# Narrow to a specific file
tyf show calculate_sum --file src/math.py

Find Symbol by Name

Searches the workspace for a symbol's definition. Supports multiple symbols in a single call. Use --fuzzy for partial/prefix matching with richer output (kind + container):

tyf find calculate_sum

# Find multiple symbols at once (results grouped by symbol)
tyf find calculate_sum multiply divide

# Narrow to a specific file (text-based search + goto_definition)
tyf find function_name --file myfile.py

# Fuzzy/prefix match (returns symbol kind + container info)
tyf find handle_ --fuzzy

Find References

# By position (exact, pipeable from list)
tyf refs -f myfile.py --line 10 --column 5

# By name
tyf refs my_function MyClass

# Mixed and piped
tyf refs file.py:10:5 my_func
... | tyf refs --stdin

Members (Class Public API)

tyf members MyClass

Call Tree

Requires ty 0.0.41 or newer.

# What does this call, transitively? (2 levels by default, max 5)
tyf calls process_order

# Who calls this? (impact analysis before an edit)
tyf calls check_inventory --in

Document Outline

tyf list src/services/user.py

Daemon Management

The daemon starts automatically on first use. Run tyf daemon --help for manual control.

Output Formats

All commands support --format (placed before the subcommand): human (default), json, csv, paths.

tyf --format json show MyClass
tyf --format csv find User --fuzzy

Architecture

CLI Command → Daemon Client (auto-connects) → Unix Socket
→ Daemon Server (5min idle timeout) → LSP Client Pool → ty LSP Server

The daemon keeps LSP connections warm: first command takes 1-2s, subsequent commands 50-100ms. See How it works for details.

Development

cargo build --release
cargo test
cargo clippy
cargo fmt --check

# Verbose logging
RUST_LOG=ty_find=debug cargo run -- find hello_world

Testing / supported ty versions

tyf does not ship ty — it drives whatever ty LSP server you have installed. ty is 0.0.x and changes frequently, so the integration suite is run in CI against a curated set of pinned, exact ty versions on every push and PR (and on dependabot ty bumps).

The tested floor is 0.0.15 and the current latest tested is 0.0.49. This is a practical baseline, not a hard limit: the suite behaves identically (same output, same ~180ms cold start) across the whole modern range — genuine capability gaps only exist in ty ≤ 0.0.5, which don't return multi-file workspace symbols (0.0.1 has no Linux glibc wheel at all).

The exact list of tested versions is the single source of truth in ci/ty-versions.json; the methodology and per-version findings are in docs/dev/TY_VERSIONS.md.

Troubleshooting

# Check ty is installed
ty --version

# Debug daemon issues
tyf daemon status
RUST_LOG=ty_find=debug tyf daemon start

# Restart daemon
tyf daemon stop && tyf daemon start

Contributing

Contributions welcome! Please open an issue to discuss major changes.

License

MIT License - see LICENSE file for details.

Credits

Built with ty - Astral's Python type checker.

Metadata

Release files for ty-find 0.5.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for ty-find 0.5.0
File Size Uploaded
ty_find-0.5.0.tar.gz 1.0 MB Details

Built distributions (wheels)

Table of built distributions (wheels) for ty-find 0.5.0
File Interpreter ABI Platform
ty_find-0.5.0-py3-none-win_amd64.whl Python 3 none Windows x86-64 Details
ty_find-0.5.0-py3-none-manylinux_2_28_x86_64.whl Python 3 none Linux glibc 2.28+ x86-64 Details
ty_find-0.5.0-py3-none-macosx_11_0_arm64.whl Python 3 none macOS 11.0+ ARM64 Details

Total release size: 4.7 MB

Release files / ty_find-0.5.0.tar.gz

Download URL ty_find-0.5.0.tar.gz
Size 1.0 MB
Tags Source
SHA-256 checksum
How to use checksums
bab7b8a107048271431853d02ac988ed65a9566cd19c3db8f801efe6ff5c1172
BLAKE2b-256 checksum
How to use checksums
c001e96f1bde06497386f9cb3a33225aa19b0a7aa07e2adef184ffe108d2468e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 6, 2026.

Transparency log

Release files / ty_find-0.5.0-py3-none-win_amd64.whl

Download URL ty_find-0.5.0-py3-none-win_amd64.whl
Size 924.2 kB
Tags Python 3 Windows x86-64
SHA-256 checksum
How to use checksums
dd2aca8b306b415a1bae6cc075201b287b95e78dd40d655b8c448e9c264b4661
BLAKE2b-256 checksum
How to use checksums
3ee640493d4edf785ed26220d7e0ce56b4e5fe27175ed728e683c72717667c67
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 6, 2026.

Transparency log

Release files / ty_find-0.5.0-py3-none-manylinux_2_28_x86_64.whl

Download URL ty_find-0.5.0-py3-none-manylinux_2_28_x86_64.whl
Size 1.4 MB
Tags Linux glibc 2.28+ x86-64 Python 3
SHA-256 checksum
How to use checksums
fed1dca32166800d391eb681050145104795418b7ac1413eb0987f88e6c4e9b7
BLAKE2b-256 checksum
How to use checksums
959d1f5f744ad14f3259a03737c03b7d9d4f936969f91314f77627e3d0c42adb
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 6, 2026.

Transparency log

Release files / ty_find-0.5.0-py3-none-macosx_11_0_arm64.whl

Download URL ty_find-0.5.0-py3-none-macosx_11_0_arm64.whl
Size 1.3 MB
Tags Python 3 macOS 11.0+ ARM64
SHA-256 checksum
How to use checksums
41b641f3c43421d1500b1c796b437772d7b49863229bfd3ce247506d9c9edd83
BLAKE2b-256 checksum
How to use checksums
01819b5d590d72393b7092b83ca575e35320295de61911e9448bba152c6fa021
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Sep 6, 2026.

Transparency log

Release history Release notifications | RSS feed

0.5.2

4 release files

0.5.1

4 release files

This release

0.5.0 This release

4 release files

0.4.6

4 release files

0.4.5

4 release files

0.4.3

4 release files

0.4.2

4 release files

0.4.1

4 release files

0.4.0

4 release files

0.3.1

4 release files

0.3.0

4 release files

0.2.2

4 release files

0.2.1

4 release files

0.2.0

4 release files

0.1.11

4 release files

0.1.10

4 release files

0.1.8

3 release files

0.1.7

3 release files

0.1.6

3 release files

0.1.5

3 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page