anishfyi

Trove — Development skill for Claude Code

Development community

A Claude Code skill that builds and maintains a personal, file-based semantic layer as you work, and reloads it every session.

How to install Trove

This entry records only its repository, not the path inside it, so there is no exact command to give. Open anishfyi/trove and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Trove does

A Claude Code skill that builds and maintains a personal, file-based semantic layer as you work, and reloads it every session.

Alternatives in Development

Create Handoff — You are tasked with writing a handoff document to hand off your work to another agent in a new session 10k ★
Obsidian Reconcile — Find and resolve contradictions in the vault — the vault maintains its own truth 277 ★
Nonstop — Autonomous work mode for Claude Code 263 ★

README

Trove

**Beyond context windows.**

Trove makes LLMs operate over repositories and knowledge bases far larger than their native context window. Not by stuffing more tokens into the prompt, but by indexing everything once and retrieving only what matters, at the right level of detail.

Live page: **https://anishfyi.com/trove/**

Two layers, one system

Layer	What it is	Where it lives
Engine	Symbol index, module/subsystem/architecture summaries, dependency graph, progressive retrieval	`.trove/` in any repo (Rust CLI)
Plugin	Personal engineering memory: decisions, gotchas, conventions	`~/.claude/trove` or `./.claude/trove` (Claude Code)

The engine answers "what is in this codebase and where?" The plugin answers "what did we decide and why?" Together they give Claude both structural repo knowledge and durable session memory.

See [VISION.md](VISION.md) for the full memory hierarchy design (L0-L6).

Quick start: engine

Requires Rust. From any repository:

git clone https://github.com/anishfyi/trove
cd trove && cargo build -p trove-cli

# Index the current repo (writes .trove/, gitignored)
cargo run -p trove-cli -- index

# Check what was indexed
cargo run -p trove-cli -- status

# Retrieve context for a task (architecture → subsystem → module → symbol)
cargo run -p trove-cli -- query "modify vendor onboarding"

Other commands: `import-historical`, `record`, `patch`. Run `cargo run -p trove-cli -- --help` for the full list.

What gets indexed

L1 Symbols: functions, classes, structs, traits, tests (Rust, Python, JS/TS, Go, Bash)
L2 Modules: per-file summaries: exports, dependencies, assumptions, side effects
L3 Subsystems: package-level clusters
L4 Architecture: repo-wide design map and data flow
L5 Historical: imported from your Claude trove entries on demand