culprit

Why an agent needs this

An agent is good at writing a patch and bad at knowing whether the patch is done. It cannot see that the helper it just changed has three other callers, or that the bug it is fixing was introduced by a refactor two years ago. culprit answers both from git history, as structured JSON it can act on rather than prose it has to interpret:

• ⚡️ Instant. No test runs, no bisect. It blames the fix's own lines back through refactors.

• 🔒 Read-only and offline. Never writes to your repo or PR. No API key, no network.

• 🔌 MCP-native. 11 tools over stdio, on the official MCP registry, plus a one-command plugin.

• 🗣️ Language-agnostic. Suspects work anywhere git does. Blast radius reads 12 language families.

Related MCP server: knownissue

Does it actually work?

Benchmarked against 50 real regressions, 25 from git and 25 from systemd, where the introducing commit is known from each fix's Fixes: trailer (author-verified ground truth). Given only the fix commit, culprit blames the removed lines to rank the commits that introduced the bug.

Deterministic and offline, on large C codebases the engine has never seen. Reproduce with python benchmarks/run.py, which clones the repos and scores every case.

Quick start

Claude Code plugin (installs the MCP server plus a skill that tells the agent when to use it):

/plugin marketplace add noordeen123/culprit
/plugin install culprit@culprit

Any MCP client (Cursor, Windsurf, VS Code, Codex CLI, Zed, Continue, Cline, Amazon Q, Goose):

{
  "mcpServers": {
    "culprit": {
      "command": "uvx",
      "args": ["--from", "culprit[mcp]", "culprit-mcp"]
    }
  }
}

CLI, no agent required:

uvx culprit                     # run from PyPI on demand
rca --verify-fix patch.diff     # exit 0 if complete, 1 otherwise

Needs Python 3.10+ and uv for the MCP server. The CLI runs on 3.9+.

The tools

verify_fix, the pre-commit gate

The agent runs this on its own diff before committing:

• verdict: complete when no untouched call site is left behind, partial when one was missed, risky when risk is high. The verdict is the completeness axis; test coverage is the separate confidence axis on risk_level.

• untouched_references: the exact files that still use the changed symbol. This is the list the agent goes and patches.

• skipped_symbols, adds_test, notes: what was too widely used to check, whether a test shipped, and what to do next.

Loop until complete, then commit. That is the whole idea.

CLI and CI

rca                            # current branch vs the configured base
rca --last                     # the latest commit only
rca --pr 16786                 # a specific PR (uses the PR's own base)
rca --trace crash.txt          # RCA from a stack trace, no fix or PR needed
rca --verify-fix patch.diff    # check a diff before committing
rca --select-tests             # print the tests to run for this change
rca --html report.html --open  # a single self-contained HTML report
rca --pr 16889 --fail-on high  # exit non-zero when QA risk is high
rca serve --repo /path         # local web UI with a base picker

CI gate: risk via exit code only, no PR comments, no writes. Copy examples/github-actions/culprit-pr.yml into .github/workflows/:

- run: pip install "culprit>=0.3.0"
- env: { GH_TOKEN: "${{ github.token }}" }
  run: rca --pr ${{ github.event.pull_request.number }} --fail-on high

HTML report

--html writes one self-contained file. No CDN, opens offline, attaches to CI. It opens with the verdict: a scored QA risk with the factors behind it, and the prime suspect with how long the bug lived.

Then it reconstructs how the bug got there. Every commit that touched the buggy line, from creation through the commit that broke it (red, with the exact diff) to the fix (green):

vs git bisect

--bisect "<cmd>" runs a real bisect as an optional confirmation layer, in a throwaway git worktree so your checkout is never touched. When the first failing commit matches the blamed suspect, the report stamps it confirmed by git bisect.

Architecture

One normalized context in, one structured JSON result out. The only non-deterministic step is the optional LLM narrative, isolated behind an adapter so the engine runs with no API key.

Two lanes, one engine. Every module writes one slice of the result, so nothing cares whether the target came from gh, the REST API, local git, or a pasted stack trace. Full module map: docs/ARCHITECTURE.md.

Configuration

The base branch resolves in order: the --base flag, then CULPRIT_BASE, then .culprit.toml (base = "origin/main"), then HEAD~1. --last forces the latest-commit view.

PR titles and labels use the GitHub CLI when present, or the unauthenticated REST API for public repos (set GITHUB_TOKEN or GITLAB_TOKEN to raise limits). Deep links cover GitHub, GitLab, Bitbucket, and Gitea. Blast radius reads imports across JS/TS, Python, Go, Java/Kotlin, Ruby, C/C++, C#, PHP, Rust, Scala, and Swift.

Contributing

pip install -e ".[dev]" && pytest

Module map and data shapes: docs/ARCHITECTURE.md. Publishing the MCP server to the registries: docs/PUBLISHING.md. MIT licensed.