pyscn is a Python code-quality analyzer that examines dead code, duplication, complexity, hotspots, architecture, dependencies, and class design, then reports which areas need attention. It is used to maintain and refactor codebases, including AI-generated code, and provides agent skills for health checks, architecture reviews, refactoring, and CI reports.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add ludo-technologies/pyscn --skill architecture-reviewgit clone --depth 1 https://github.com/ludo-technologies/pyscnWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/ludo-technologies/pyscn/architecture-review)<a href="https://agentmods.dev/skills/ludo-technologies/pyscn/architecture-review"><img src="https://agentmods.dev/badge/skills/ludo-technologies/pyscn/architecture-review/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ludo-technologies/pyscn/architecture-review"><img src="https://agentmods.dev/badge/skills/ludo-technologies/pyscn/architecture-review.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 6 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium MCP Rug Pull · line 8 uvx/uv tool run commands without ==version create a rug-pull risk.Fix: Pin the version: uvx package-name==1.2.3
- medium MCP Rug Pull · line 14 uvx/uv tool run commands without ==version create a rug-pull risk.Fix: Pin the version: uvx package-name==1.2.3
- medium MCP Rug Pull · line 15 uvx/uv tool run commands without ==version create a rug-pull risk.Fix: Pin the version: uvx package-name==1.2.3
- medium MCP Rug Pull · line 16 uvx/uv tool run commands without ==version create a rug-pull risk.Fix: Pin the version: uvx package-name==1.2.3
- medium MCP Rug Pull · line 17 uvx/uv tool run commands without ==version create a rug-pull risk.Fix: Pin the version: uvx package-name==1.2.3
- medium MCP Rug Pull · line 18 uvx/uv tool run commands without ==version create a rug-pull risk.Fix: Pin the version: uvx package-name==1.2.3
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00053 | $0.00678 |
| Opus 5 | $0.00026 | $0.00339 |
| Sonnet 5 | $0.00011 | $0.00136 |
| Haiku 4.5 | $0.00005 | $0.00068 |
Grade A, and why
architecture-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 43 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Python Architecture Review with pyscn
Run the pyscn CLI to understand module structure and coupling. No install needed: uvx pyscn@latest <command>.
Commands
| User Request | Command |
|---|---|
| "Check class coupling" | uvx pyscn@latest analyze --select cbo <path> |
| "Find circular dependencies" | uvx pyscn@latest analyze --select deps <path> |
| "Check class cohesion" | uvx pyscn@latest analyze --select lcom <path> |
| "Which files should I review together?" | uvx pyscn@latest analyze --select communities --json <path> |
| "Map the module architecture" | uvx pyscn@latest analyze --select communities --json <path> |
Useful flags: --min-cbo <n> (minimum CBO to report), --json (detailed report file). Report files are NOT written to stdout; they go to .pyscn/reports/ and the path is printed on completion.
Module Communities (context map for agents)
Module community detection is opt-in: run with --select communities --json and read the report file.
The JSON report includes community_analysis.community_context_map, a compact, deterministic map telling you which modules to inspect together:
bundles[]: clusters of modules that change together. Each hasmodules,packages,risk_level, the cluster'sbridge_modules, an optionalsuggested_review_scope(path prefix to focus a review on), and a one-linesummary. Large bundles cap the module list with a... +N moremarker (module_countalways holds the true total).bridge_modules[]: modules that couple two or more communities (module, the communities itconnects, and areason). Widen the review scope to include these when touching a bundle they connect.
Use the map to scope reviews and refactors: load all modules in a bundle together, and pull in connected bridge modules before changing cluster boundaries.
Interpreting Coupling Results
- High CBO classes depend on many others; changes ripple widely. Suggest interface extraction or dependency inversion.
- High LCOM4 classes bundle unrelated responsibilities; suggest splitting along the disconnected method groups.
- Dependency cycles are the highest-priority architectural issue; name the modules in each cycle and the weakest edge to break.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 43 lines · 53 tokens per session scan A 5ce5f826bb02
architecture-review is a skill published in the GitHub repository ludo-technologies/pyscn (1,052 stars, last pushed today), licensed MIT. It adds 53 tokens to every session and 678 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
code-quality
Python code-quality anti-patterns and review checks: exception-hierarchy correctness, singleton identity comparison, narrow exception handling, wildcard-import avoidance, magic-number naming, and dead-local removal. Use when reviewing or self-reviewing Python code for correctness and readability defects that linters…
remember
Review the current conversation and capture valuable knowledge — best practices, coding conventions, architecture decisions, workflows, and user feedback — into persistent memory (AGENTS.md) or reusable skills. Use when the user says: (1) remember this, (2) save what we learned, (3) update memory, (4) capture…
code-review
Perform a structured code review of changes, checking for correctness, style, tests, and potential issues.
code-review
Code review assistance with linting, style checking, and best practices.
red-team-review
Unified adversarial review: v4.3 Strategic Matrix (MTA-004). 7-phase framework: Priors → Rubric → Adversarial Lenses → SWOT/TOWS → MCDA Decision Engine → Blind Spot/Kill Switch → Executive Summary. Absorbs: bias-detector.
decision-journal
Unified decision lifecycle: Pre-decision logging, post-decision review, failure classification, and calibration tracking. Absorbs: post-mortem-engine.