nomograph-ainomograph-ai/rune/CLAUDE.md

rune

CLAUDE.md0 starsChanged 7 months agoMITView on GitHub

Registry manager for AI coding agent skills, agents, and rules. Syncs markdown files from git-based registries into .claude/skills/, .claude/agents/, and .claude/rules/. Tracks drift via content-hash lockfile and pedigree frontmatter. CI uses nomograph/pipeline/rust-cli@v2.4.6 with cargo-deny for advisories and license audit. audit_allow_failure: false. Lib + binary crate. lib.rs re-exports all modules for testing. - ArtifactType -- enum: Skill, Agent, Rule. Each has section name, default directory, and display name. Agents use .claude/agents/ (Anthropic's directory name; their docs say "subagents"). - Manifest --…

In this repository2 other files it gives its agents.
- resourcesSkill
- runeSkill

The file

We split this CLAUDE.md into the rules it gives an agent

2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.

≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Very specific:
All errors use anyhow::Result with .context() or bail!().
line 55
Very specific:
import/upstream/diff/update are skill-only (use pedigree metadata).
line 60

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

CLAUDE.md

rune

Registry manager for AI coding agent skills, agents, and rules. Syncs markdown files from git-based registries into .claude/skills/, .claude/agents/, and .claude/rules/. Tracks drift via content-hash lockfile and pedigree frontmatter.

Build

cargo build --release       # binary at target/release/rune
cargo test                  # 22 unit + 40 integration tests
cargo clippy -- -D warnings # lint (also enforced at crate level)

CI uses nomograph/pipeline/rust-cli@v2.4.6 with cargo-deny for advisories and license audit. audit_allow_failure: false.

Architecture

Lib + binary crate. lib.rs re-exports all modules for testing.

Module	Responsibility
`main.rs`	CLI (clap derive), command dispatch, `--type` flag parsing
`commands/mod.rs`	SkillStatus enum, resolve_registry helpers
`commands/check.rs`	Drift checking for all three types
`commands/sync.rs`	Sync engine, AGENTS.md generation, agent symlink
`commands/crud.rs`	add, push, remove (type-aware)
`commands/info.rs`	ls, status, doctor, clean, audit
`commands/upstream.rs`	browse, import, upstream, diff, update (skill-only)
`registry.rs`	Git/archive registry operations, hashing, file locking
`pedigree.rs`	YAML frontmatter parsing/writing, URL slug extraction
`manifest.rs`	ArtifactType enum, per-project rune.toml, path overrides
`config.rs`	Global config, registry definitions, artifact resolution
`setup.rs`	One-time setup: config creation, Claude Code hook install
`lockfile.rs`	Per-project rune.lock: content hashes for drift detection
`color.rs`	Terminal color support (ANSI, respects TTY)

Key types

ArtifactType -- enum: Skill, Agent, Rule. Each has section name, default directory, and display name. Agents use .claude/agents/ (Anthropic's directory name; their docs say "subagents").
Manifest -- per-project .claude/rune.toml. Sections: [skills], [agents], [rules], optional [paths] for overrides.
Lockfile -- per-project .claude/rune.lock. Matching sections.
Pedigree -- YAML frontmatter tracking origin, import date, upstream commit, modification status. Skill-only concept.
Config -- global ~/.config/rune/config.toml. Registry definitions.
SkillStatus -- drift check result: Current, Drifted, Missing, etc.

Conventions

All errors use anyhow::Result with .context() or bail!().
#![deny(warnings, clippy::all)] enforced in both lib.rs and main.rs.
Names validated against path traversal via validate_name().
Registry pulls use file locking (fs2) for concurrent safety.
Registries cached in ~/.cache/rune/registries/.
import/upstream/diff/update are skill-only (use pedigree metadata).
Global state (offline, dry_run) via AtomicBool statics in registry.rs.
Types kept as SkillEntry and LockedSkill for serde backward compat.

Testing

62 tests: 11 unit (pedigree), 11 duplicated in bin, 40 integration.

Integration tests cover: manifest/lockfile roundtrips for all types, typed and legacy registry layouts, artifact path resolution, find_type collision warnings, path overrides, and all original skill-focused tests.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

It worked
Partly
It didn't

What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free