Active coding discipline enforcer based on Karpathy's 4 principles: surface assumptions, keep it simple, make surgical changes, define verifiable goals. Ships 4 Python tools (complexity_checker, diff_surgeon, assumption_linter, goal_verifier), a review agent, /karpathy-check slash command, and a pre-commit hook. All tools stdlib-only.
- Stars
- 29,893
- Version
- v2
- Copies
- 0
Overview
Waxell rates it high risk: 14 findings, highest severity high.
- Highest severity
- high
- Findings
- 14
- What ran
- Static analysis + deep scan
A clean result means nothing alarming in what is published, not a guarantee. How the checks work
We don't republish the plugin's instructions. Read them on GitHub at the reviewed commit.
Security
Security scan
Checked 2 days ago
Result
High risk Waxell rates it high risk: 14 findings, highest severity high.
Checked with static analysis + deep scan: servers it launches, packages it installs at runtime, scripts piped to a shell and bundled binaries. The deep scan also read the instructions for prompt injection and data exfiltration. Nothing was run. How the checks work
Findings (14)
high · 1
- high
Tool Misuse: Tool Parameter Abuse
agents/karpathy-reviewer.md:8 · skillspector:TM1
medium · 13
- medium
Excessive Agency: Autonomous Decision Making
README.md:82 · skillspector:EA2
- medium
Excessive Agency: Autonomous Decision Making
skills/karpathy-coder/expected_outputs/goal_verifier.json:17 · skillspector:EA2
- medium
Excessive Agency: Autonomous Decision Making
skills/karpathy-coder/scripts/assumption_linter.py:104 · skillspector:EA2
- medium
Excessive Agency: Autonomous Decision Making
skills/karpathy-coder/scripts/goal_verifier.py:24 · skillspector:EA2
- medium
Excessive Agency: Autonomous Decision Making
skills/karpathy-coder/scripts/goal_verifier.py:156 · skillspector:EA2
- medium
Agent Snooping: Skill Enumeration
commands/karpathy-check.md:47 · skillspector:AS3
- medium
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and pub
hooks/karpathy-gate.sh:8 · skillspector:RP1
- medium
npx commands without a version suffix (e.g. @1.0.0) create a rug-pull risk if the upstream server is compromised and pub
skills/karpathy-coder/references/enforcement-patterns.md:37 · skillspector:RP1
- medium
Excessive Agency: Autonomous Decision Making
skills/karpathy-coder/SKILL.md:17 · skillspector:EA2
- medium
Excessive Agency: Autonomous Decision Making
skills/karpathy-coder/references/karpathy-principles.md:11 · skillspector:EA2
- medium
Excessive Agency: Autonomous Decision Making
skills/karpathy-coder/references/karpathy-principles.md:15 · skillspector:EA2
- medium
Excessive Agency: Autonomous Decision Making
skills/karpathy-coder/scripts/assumption_linter.py:61 · skillspector:EA2
- medium
Dangerous Code Execution: subprocess module call
skills/karpathy-coder/scripts/diff_surgeon.py:54 · skillspector:AST4
Versions
Version history
Each distinct content we have seen, newest first, with the verdict it got at the time.
- v2 (2.14.9)2026-11-162 changed High riskgot riskier
- v1 (2.14.6)2026-11-03 No issues
Files
Files
17 files, 63 KB in total. Names only; open them on GitHub.
- .claude-plugin/plugin.json
- README.md
- agents/karpathy-reviewer.md
- commands/karpathy-check.md
- hooks/karpathy-gate.sh
- skills/karpathy-coder/SKILL.md
- skills/karpathy-coder/expected_outputs/assumption_linter.json
- skills/karpathy-coder/expected_outputs/complexity_checker.json
- skills/karpathy-coder/expected_outputs/diff_surgeon.json
- skills/karpathy-coder/expected_outputs/goal_verifier.json
- skills/karpathy-coder/references/anti-patterns.md
- skills/karpathy-coder/references/enforcement-patterns.md
- skills/karpathy-coder/references/karpathy-principles.md
- skills/karpathy-coder/scripts/assumption_linter.py
- skills/karpathy-coder/scripts/complexity_checker.py
- skills/karpathy-coder/scripts/diff_surgeon.py
- skills/karpathy-coder/scripts/goal_verifier.py
See and control which skills your people use
Inside your organization, Waxell scans each skill and plugin before anyone installs it, approves what passes and publishes it to a private marketplace, and puts every MCP tool call behind one governed URL.
30 minutes. No commitment.
SOC 2 Type II attested · ISO/IEC 27001:2022 certified · HIPAA with a BAA · aligned to the EU AI Act and NIST AI RMF
More engineering skills and plugins
pr-review-expertalirezarezvaniFile-by-file pull request review for bugs, security and quality, with a merge verdict. Load whenever the user asks to review, look over or… High riskEngineeringSecurity30k
cmux-architecturemanaflow-aicmux package architecture, refactor layering, dependency inversion, file organization, DocC documentation, package design discipline,… No issuesEngineeringDesign29k
cmux-backendmanaflow-aiBackend TypeScript and Cloud VM development rules for cmux. Use when editing web/app/api, web/services, backend scripts, Cloud VM… No issuesEngineeringImages, video & audio29k
cmux-capturemanaflow-aiScreenshot or record a cmux window from the CLI, for a pull request, a bug report, or to see what the app actually looks like right now.… No issuesEngineeringDocuments29k