Skills Explorer Add Skills

genotoxic

An agent skill by trailofbits, from trailofbits/skills. Tags: analysis, code-quality, fuzzing, testing.

What it does

Graph-informed mutation testing triage. Parses codebases with Trailmark, runs mutation testing and necessist, then uses survived mutants, unnecessary test statements, and call graph data to identify false positives, missing test coverage, and fuzzing targets. Use when triaging survived mutants, analyzing mutation testing results, identifying test gaps, finding fuzzing targets from weak tests, running mutation frameworks (including circomvent and cairo-mutants), or using necessist.

Install

With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:

npx skills add trailofbits/skills --skill genotoxic

Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):

git clone --depth 1 https://github.com/trailofbits/skills
cp -r skills/plugins/trailmark/skills/genotoxic ~/.claude/skills/genotoxic

Safety box score

Grade C, moderate risk: 39 of 100 risk points, rated 2026-10-20. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.

This skill analyzes a codebase’s tests and graph data, but it also runs mutation frameworks and project test commands, which can execute arbitrary repository code. The main risks are broad code execution during analysis, plus installing external tooling from the network and, in some paths, using sudo to install system-wide components.

Category levels, from 0 (none) to 3 (high)
CategoryLevelWhy
Code & shell execution3 (high)It explicitly runs test and mutation commands against the target project (`uv run mutmut run`, `mull-runner ... ./build/tests`, `slither-mutate . --test-cmd "forge test"`), which means arbitrary repository code can execute during normal use.
Network access2 (moderate)The skill installs and fetches tools from external sources, including `uv add --dev pytest-gremlins`, `go install ...@latest`, `gh release download`, and `curl -1sLf ... | sudo -E bash`.
Filesystem reach1 (low)Primary writes are scoped to the workspace, such as `GENOTOXIC_REPORT.md`, but the workflow also creates tool output directories like `mutation_campaign/` and `mull-report/`.
Secrets & credentials0 (none)The files do not ask for API tokens, passwords, browser sessions, or credential-store access, and they do not reference `.env` or keychain reads.
Privilege & persistence2 (moderate)Some installation paths require elevated system changes, e.g. `sudo apt-get install clang-<version>`, `sudo mkdir -p /usr/local/bin /usr/local/lib`, and `sudo cp ... /usr/local/bin/`.
Instruction hijack surface0 (none)There is no prompt-injection-style content telling the agent to hide actions, ignore higher-priority instructions, or follow remote instructions beyond normal install guidance.
Irreversible actions0 (none)The skill does not deploy, publish, send messages, pay, book, or delete external resources; it only triages local code and writes a report.
Transparency2 (moderate)Several setup paths rely on unpinned or external content not shipped with the skill, such as `go install ...@latest`, `gh release download`, and `git clone` for `circomvent` and `cairo-mutants`, plus remote installer scripts.

Before installing

Rated 2026-10-20 by an automated review of 6 files from the skill's repository.

Source

Repository
trailofbits/skills (all skills from this repository)
Path
plugins/trailmark/skills/genotoxic/SKILL.md
Branch
main
Collection
trailmark
Updated
2026-09-19