AIttributor
https://github.com/block/aittributor- Category
- Developer Tools
- Rank
- No. 1653Tools index
Previous survey · No. 1658 ·
- Pricing
- Open Source
- Type
- TOOL
- Builder
- block
- GitHub
- 6 stars
- Latest release
- v0.8.0
- Date
About
A prepare-commit-msg git hook that adds AI agent attribution to your commits.
What it does
AIttributor observes the commit environment, nearby processes, and recent agent session records associated with the repository. It selects the first detected agent and adds that agent’s official email as a co-author trailer. Branch matching helps narrow session-based detection when both sides expose a branch.
Why it's ranked here
This is a focused, pragmatic tool for teams that want AI involvement recorded in ordinary Git history. Multiple detection strategies make it useful beyond agent-initiated commits, while duplicate checks and a hard timeout reduce disruption. Its attribution remains heuristic, so the record is useful evidence rather than definitive provenance.
What's good
Detection covers environment signals, process ancestry, related process trees, and recent session records. Repository and branch checks constrain the broader fallback. Existing trailers with the same email are preserved without duplication. Failure or slow detection does not block the commit beyond the one-second timeout.
Tradeoffs
Session history can over-attribute when an agent only performed research, and manual commits are less reliable than agent-initiated commits. Session fallback supports only agents with readable state files. Branch data may be stale or absent. When several agents appear, only the first match receives credit.
How to use it well
Use it in teams that accept Git trailers as a lightweight disclosure record and commonly let coding agents initiate commits. Connect it through Lefthook or a repository hook, then treat results as auditable hints. It does not provide exact authorship provenance, contribution measurement, or complete multi-agent attribution.
Technical notes+
Cargo.toml defines a Rust 2024 binary using clap, sysinfo, and whoami. src/main.rs runs breadcrumb discovery on a separate thread, combines it with environment and process-tree detection, chooses the first result, and enforces a one-second timeout. src/agent.rs centralizes supported process names, environment markers, emails, and breadcrumb metadata. src/breadcrumbs.rs scans recent JSONL session files, reads at most five lines per file, and checks repository paths plus branches when available. src/git.rs uses git symbolic-ref for branch discovery and git interpret-trailers for in-place message updates. install.sh downloads compressed release binaries for supported operating systems and architectures.
Observed
- License
- Apache License 2.0
- Primary language
- Rust, using the 2024 edition
- Interface
- Command-line binary designed for Git prepare-commit-msg hook integration
- Installation
- Shell installer downloads a compressed prebuilt release binary; custom installation directories are supported
- Platform support
- macOS and Linux on x86_64 or aarch64
- Hook integration
- Documented configurations cover Lefthook and direct Git hook symlinks
- Test structure
- Unit tests are embedded in the Rust source modules shown
Read from README.md, Cargo.toml, src/main.rs, src/git.rs, src/agent.rs, src/breadcrumbs.rs, LICENSE, Justfile, AGENTS.md, CLAUDE.md, CODEOWNERS, install.sh, lefthook.yml, rustfmt.toml, GOVERNANCE.md.
What it can do
Add AI agent attribution to git commits automatically
Git commit being prepared → Commit message with AI agent attribution added
Install as a git prepare-commit-msg hook
Git repository → Configured git hook for automatic attribution
Detect when AI assistance was used in code changes
Code changes in commit → Attribution metadata identifying AI involvement
Modify commit messages before they are finalized
Original commit message → Enhanced commit message with AI attribution tags
Track AI tool usage in version control history
Series of commits with AI assistance → Git history with documented AI contributions
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.