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 beyondnetcode/evolith_arch32 --skill architecture-governancegit clone --depth 1 https://github.com/beyondnetcode/evolith_arch32Wrote 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/beyondnetcode/evolith_arch32/architecture-governance)<a href="https://agentmods.dev/skills/beyondnetcode/evolith_arch32/architecture-governance"><img src="https://agentmods.dev/badge/skills/beyondnetcode/evolith_arch32/architecture-governance/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/beyondnetcode/evolith_arch32/architecture-governance"><img src="https://agentmods.dev/badge/skills/beyondnetcode/evolith_arch32/architecture-governance.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00076 | $0.00811 |
| Opus 5.5 | $0.00030 | $0.00324 |
| Sonnet 5.5 | $0.00015 | $0.00162 |
| Haiku 4.5 | $0.00008 | $0.00081 |
Grade A, and why
architecture-governance 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 1mo 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 — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Architecture governance with Evolith
Validate a repository against an executable rule corpus, and report the verdict together with its coverage.
The one thing that matters
evolith validate returns four counts: checked, skipped, errored, total.
Never report a run as passing without stating skipped and errored. Those rules were
not evaluated. Their result is UNKNOWN, not passed. A run with 133 checked and 26 skipped is
not "133 rules passed" -- it is "133 evaluated, 26 unknown", and a blocking rule among the
unknown is a failure.
This is the whole reason the tool exists. Every other linter collapses "passed" and "never ran" into the same green, so their coverage and their compliance look identical.
How to run it
The repository must be an Evolith satellite: it needs an evolith.yaml. If there is none,
say so rather than initializing one uninvited -- evolith init writes to the user's repo.
evolith validate --engine opa --format json
If the CLI is not installed, use npx -y @beyondnet/evolith-cli validate --engine opa --format json.
Prefer --format json: it prints exactly one envelope on stdout and never prompts.
Reading the exit code
It is a taxonomy, not a boolean. Read it before reading the report.
| Code | Meaning | What to tell the user |
|---|---|---|
0 |
passed | Report the verdict, and still state the skipped count |
1 |
the tool failed | No verdict exists. Do not report compliance either way |
2 |
the gate blocked | A real verdict, and it says no |
3 |
invalid invocation | Nothing was evaluated. Fix the command |
1 and 3 are not weaker forms of 2. If you report them as "some issues found", you have
told the user their repository was examined when it was not.
Reading the report
From the JSON envelope's data:
status--passed,warningorfailedrulesChecked/rulesSkipped/rulesErrored/rulesTotalskippedRuleIds-- name them when the user asks what was not coveredissues[]-- each withblocking,ruleId,severity,fileselection-- which rulesets ran.source: core-defaultmeans the whole corpus; a caller selection means a narrower claim, and you must say so
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.
- 1mo ago First seen · 78 lines · 76 tokens per session scan A cc807656b56b
architecture-governance is a skill published in the GitHub repository beyondnetcode/evolith_arch32 (2 stars, last pushed 17d ago), licensed MIT. It adds 76 tokens to every session and 811 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-08-31.
Other skills, from other repositories
codebase-review
Review an entire codebase for architecture, engineering health, and exploitable risk; generate a prioritized remediation plan, an evidence-anchored system knowledge document, or both.
review-system-design
Review a system design someone else provided - screenshot, drawio, Mermaid, slides, doc, or IaC - by extracting it into a confirmed fact sheet, then scoring it on the nine axes.
code-review
Run an AI-assisted PR code review using multi-layer lenses with confidence scoring.
review-ticket
Review a ticket or PR through focused specialist lenses: scope, architecture, security, tests, AC coverage, and PR metadata.
common-feedback-reporter
Pre-write audit for skill violations: checks planned code against loaded skill anti-patterns before any file write. Use when writing Flutter/Dart/TS code or editing SKILL.md files with active project skills. Load as composite; on auto-fixed violation, also load +common/common-learning-log.
common-code-review
Conduct high-quality, persona-driven code reviews. Use when reviewing PRs, critiquing code quality, or analyzing changes for team feedback.