Skip to main content

Spec-Check

Did the AI actually build what you asked for?

One command. Checks your spec against the built code. Shows what's done, what's missing.

$ spec-check --spec "login, signup, dashboard, tasks, dark mode" --dir ./my-app

Built by NOUMENON — AI agents that debate, evolve, and build.

PyPI License: MIT Python 3.10+

Why This Exists

Every developer using Copilot, Cursor, or Claude has this problem: "I asked for 6 pages and it built 3." spec-check catches what's missing.

Install

pip install spec-check

# Optional: LLM-enhanced matching
pip install spec-check[llm]

Usage

# Check from text spec
spec-check --spec "login page, signup page, dashboard, task list, dark mode" --dir ./my-app

# Check from spec file
spec-check --spec-file spec.md --dir ./my-app

# Auto-detect from README
spec-check --dir ./my-app --auto

# JSON output for CI/CD
spec-check --spec "..." --dir . --format json

# Strict mode (exit 1 if < 100%)
spec-check --spec "..." --dir . --strict

Output

  SPEC-CHECK v1.0.0

  Scanning: ./my-app (47 files)
  Spec features: 7

  Feature              Status    Evidence
  Login page           FOUND     src/pages/login.tsx
  Signup page          FOUND     src/pages/signup.tsx
  Dashboard            FOUND     src/pages/dashboard.tsx
  Task list            FOUND     src/components/Tasks.tsx
  Task creation form   MISSING
  Dark mode            PARTIAL   tailwind.config has dark
  Mobile responsive    FOUND     Breakpoints in CSS

  Completeness: ██████████████░░░░░░ 71% (5/7 features built)

  Missing: task creation form
  Partial: dark mode (config exists, not applied to all components)

How It Works

  1. Parse your spec into individual features
  2. Generate search keywords for each feature (with synonyms)
  3. Search all project files for those keywords (filenames + contents)
  4. Score each feature: FOUND (2+ matches), PARTIAL (1 match), MISSING (0)
  5. Calculate completeness percentage

CI/CD Integration

name: Spec Check
on: [push, pull_request]
jobs:
  verify:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - run: pip install spec-check
      - run: spec-check --spec-file SPEC.md --dir . --strict

Contributing

Ideas for better feature matching? Open a PR.

License

MIT


Part of the NOUMENON ecosystem. NOUMENON doesn't just check specs — it builds 100% of them. 16 AI agents debate, then build, then verify. Every time.

Metadata

Release files for ai-spec-check 1.0.0

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

Source distribution (sdist)

Source distribution for ai-spec-check 1.0.0
File Size Uploaded
ai_spec_check-1.0.0.tar.gz 10.0 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for ai-spec-check 1.0.0
File Interpreter ABI Platform
ai_spec_check-1.0.0-py3-none-any.whl Python 3 none any Details

Total release size: 21.7 kB

Release files / ai_spec_check-1.0.0.tar.gz

Download URL ai_spec_check-1.0.0.tar.gz
Size 10.0 kB
Tags Source
SHA-256 checksum
How to use checksums
69f310de0cd82782eb6d6a2284d2ba4d992bcd001b9618ea75a4501e8a7b99f4
BLAKE2b-256 checksum
How to use checksums
13f1d0e0ed78bb3d96ba873329a753703e1f7b0744f7b5fae3cd9fc97f7a87d9
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.13.7

Release files / ai_spec_check-1.0.0-py3-none-any.whl

Download URL ai_spec_check-1.0.0-py3-none-any.whl
Size 11.7 kB
Tags Python 3
SHA-256 checksum
How to use checksums
92f1005da08d433e36e2e6a3f03c9b315bbda571c961533f4902ee4911a0bd6d
BLAKE2b-256 checksum
How to use checksums
4ae9e9819a6d1308b239af34f78086a11f2d99e78a8908aa205cc14d620f88db
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/6.2.0 CPython/3.13.7

Release history Release notifications | RSS feed

This release

1.0.0 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