Skip to main content

agentlint

Lint, score, and optimize your AI agent configuration files.

agentlint analyzes CLAUDE.md, AGENTS.md, Cursor rules, and other AI agent configs — and tells you exactly what's wrong and how to fix it.

Install

pip install agentlint

Usage

# Lint a file and show issues
agentlint lint CLAUDE.md

# Score your config from 0-100
agentlint score CLAUDE.md

# Generate an optimized version
agentlint optimize CLAUDE.md          # saves to CLAUDE.optimized.md
agentlint optimize CLAUDE.md --ai     # uses Claude API for smarter suggestions

# Show diff of what optimize would change
agentlint diff CLAUDE.md

# Scan entire directory for all config files
agentlint scan .

Lint Rules

Code Severity Description
TOO_LONG ❌ Error File exceeds 80 lines — Claude starts ignoring instructions
TOKEN_WASTE ⚠️ Warning Personality instructions waste tokens with little effect
DUPLICATE_RULE ⚠️ Warning Same instruction repeated more than once
CONTRADICTION ❌ Error Conflicting instructions cause unpredictable behavior
MISSING_COMMANDS ❌ Error No shell commands defined — agent can't verify its work
MISSING_STACK ❌ Error No tech stack defined — agent may make wrong assumptions
AT_MENTION_ABUSE ⚠️ Warning Excessive @file refs embed large files every session
NO_VERIFY_STEP ❌ Error No instructions for verifying changes work
VAGUE_INSTRUCTION ⚠️ Warning Instruction too vague to be reliably actionable
BLOAT_DETECTED ⚠️ Warning Lengthy procedures belong in skill files, not base config
MISSING_NEVER ⚠️ Warning No prohibition rules — negative constraints are highly effective
INSTRUCTION_LIMIT ⚠️ Warning Estimated 150+ instructions — models lose track

Scoring

Category Points What it measures
Length & Density 25 File size, no token waste, no duplicates
Specificity 25 Concrete rules, no vague/contradicting instructions
Structure 25 Headers, bullets, code blocks, no bloat
Completeness 25 Stack, commands, verify step, never rules

License

MIT

Metadata

Release files for agentlint-dev 0.1.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for agentlint-dev 0.1.0
File Size Uploaded
agentlint_dev-0.1.0.tar.gz 40.1 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for agentlint-dev 0.1.0
File Interpreter ABI Platform
agentlint_dev-0.1.0-py3-none-any.whl Python 3 none any Details

Total release size: 57.8 kB

Release files / agentlint_dev-0.1.0.tar.gz

Download URL agentlint_dev-0.1.0.tar.gz
Size 40.1 kB
Tags Source
SHA-256 checksum
How to use checksums
26e59260072129ef5e23c1c49d251cdd3a24b4c09149c7c02ba4bb6b0d990209
BLAKE2b-256 checksum
How to use checksums
327d642c07ff35138531c0cd2a1c6c41f04d821e449295d4e68df9522e05b040
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.9.12

Release files / agentlint_dev-0.1.0-py3-none-any.whl

Download URL agentlint_dev-0.1.0-py3-none-any.whl
Size 17.7 kB
Tags Python 3
SHA-256 checksum
How to use checksums
bd29c4da7b9ee2a17f93d58cfe43270c91cdefbfb879fb335f1869edf5951c2b
BLAKE2b-256 checksum
How to use checksums
1f3e702afc2ef2b298cfad9b4849710f5d9e841c9359384b3d3b4c56a572cdaa
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.9.12

Release history Release notifications | RSS feed

This release

0.1.0 This release

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page