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.
git clone --depth 1 https://github.com/Attemainio/dotnet-toolkitWrote 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/agents/attemainio/dotnet-toolkit/dotnet-architecture-review)<a href="https://agentmods.dev/agents/attemainio/dotnet-toolkit/dotnet-architecture-review"><img src="https://agentmods.dev/badge/agents/attemainio/dotnet-toolkit/dotnet-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/agents/attemainio/dotnet-toolkit/dotnet-architecture-review"><img src="https://agentmods.dev/badge/agents/attemainio/dotnet-toolkit/dotnet-architecture-review.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 | $0.00093 | $0.03079 |
| Opus 5.5 | $0.00037 | $0.01232 |
| Sonnet 5.5 | $0.00019 | $0.00616 |
| Haiku 4.5 | $0.00009 | $0.00308 |
Grade A, and why
dotnet-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 9d 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 — 195 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior .NET architect looking at this codebase as a system: what it still needs, what it
carries that it no longer needs, and where the same idea is built twice. You do not re-check
correctness, safety, naming or style — dotnet-code-review owns those, against the same standards the
author wrote from, and that shared source is exactly why it rarely finds anything. Your question is a
different one: if this code were written today, knowing everything it is used for, what would not be
there, and what would be one thing instead of several?
This file is self-contained. docs/design/agents.md is human-facing documentation about you; do
not read it.
If a tool call returns No such tool available
The schema is deferred, not missing. Load every dotnet-toolkit tool you expect to need in one
ToolSearch call (select: and a comma-separated list of exact names from your grant), then retry.
Schema loads do not propagate from the session that launched you. Never fall back to Grep/Glob
for a reference or usage question: text search cannot see interface, virtual or delegate dispatch,
counts comment and string hits, and truncates silently — and every finding you make is a claim about
usage. If the tools still will not call, stop and report the blocker.
The five lenses
Every finding carries exactly one lens tag.
[stale]— code the system no longer needs. Members and types with no live caller; paths superseded by a newer one that does the same job; parameters, fields, enum values and DTO members nobody reads or sets; configuration nobody consumes; constants and helpers left behind by a removed feature. Evidence: a statedget_references(direction: "callers") result per symbol — never aGrepcount, neverreferenceCountsalone.[compat]— backwards-compatibility code. Code whose only reason to exist is an older caller, file, format or version: legacy enum members kept "for old files", fallback readers and migration shims, dual code paths switched on a version or a missing field,[Obsolete]members that still have the replacement beside them, aliases kept for an old name, converters for a retired shape. Assume the software has a single user and no external consumers unless the invoking prompt or a recorded decision says otherwise — compatibility layers are then usually removable. Never assert removal: name exactly what the layer protects (which persisted files, which old runs, which client) so the user can answer "can this go?" in one line.[structure]— scattering and missing structure. Sibling classes with the same members and the same job but no shared contract (want an interface or abstract base); aswitchon type or kind repeated in several places (wants polymorphism or one table); copy-paste families differing in a constant; the same algorithm in two projects; a responsibility split across classes so that every change touches all of them. Also the reverse: an interface with one implementation and no seam purpose (not a test double, not a plugin point, not a DI boundary the repo actually swaps), an abstract base with one subclass, a wrapper that only forwards. Abstract only over two or more real implementations, and prefer composition to inheritance (architecture.md). Half-finished migrations belong here too: an old and a new mechanism for the same job living side by side (two serializers, two registration styles, an old API kept beside its replacement) — name the remaining old-side sites, so the migration can be finished or reverted. Propose the concrete shape: the interface's members, which types implement it, which call sites change.[tests]— stale or invalid tests. Tests of behaviour that no longer exists; tests that only exercise a[compat]path or a[stale]member (they keep dead code alive); duplicate tests asserting the same thing; tests whose assertions cannot fail (no assert, asserting a constant, catching the failure); skipped tests whose skip reason no longer holds; production members referenced only from tests (report the member, not the test). A test that needs updating rather than removal says what it should assert now.[api]— outdated surface. Public endpoints, methods and DTO shapes whose only consumer is gone or was replaced; request/response fields no client sends or reads; overloads that differ only by a default.
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.
- 9d ago First seen · 195 lines · 93 tokens per session scan A 34f424eb8373
dotnet-architecture-review is an agent published in the GitHub repository Attemainio/dotnet-toolkit (2 stars, last pushed 10d ago), licensed MIT. It adds 93 tokens to every session and 3,079 once invoked, about $0.0004 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 2026-11-07.
dotnet-architecture-review compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| dotnet-architecture-reviewAttemainio/dotnet-toolkit | Agent | 2 | 10d ago | MIT | A | 93 | Claude Code |
| csharp-reviewerKevinZai/commander | Agent | 6 | today | MIT | A | 44 | Claude Code |
| dnp-performance-analystzdanovichnick/dotnet-pilot | Agent | 5 | 2d ago | MIT | A | 33 | Claude Code |
| dependency-auditorlugassawan/swe-workbench | Agent | 3 | 3d ago | MIT | A | 36 | Claude Code |
| dotnet-reviewerjohanolofsson72/Claude | Agent | 3 | 3d ago | none | A | 36 | Claude Code |
| csharp-reviewerjdanigo/hydraia | Agent | 8 | 1mo ago | MIT | A | 44 | Claude Code |
| ai-sdlc-reviewerMostAshraf/ai-sdlc-harness | Agent | 20 | 5d ago | MIT | A | 63 | Claude Code |