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/mickeyyaya/evolve-loopWrote 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/mickeyyaya/evolve-loop/go-reviewer)<a href="https://agentmods.dev/agents/mickeyyaya/evolve-loop/go-reviewer"><img src="https://agentmods.dev/badge/agents/mickeyyaya/evolve-loop/go-reviewer/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/mickeyyaya/evolve-loop/go-reviewer"><img src="https://agentmods.dev/badge/agents/mickeyyaya/evolve-loop/go-reviewer.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 | $0.00053 | $0.00884 |
| Opus 5.5 | $0.00021 | $0.00354 |
| Sonnet 5.5 | $0.00011 | $0.00177 |
| Haiku 4.5 | $0.00005 | $0.00088 |
Grade C, and why
go-reviewer scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
<!-- GENERATED:instructions BEGIN from agents/evolve-diff-go-reviewer.md by `evolve skills generate`; edit the source, not this region --> This is a copy
95% identical to go-reviewer — 114 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior Go reviewer. You review the DIFF, not the package, and you are cheap by design: a 150-line diff must cost well under 40K tokens.
Scope (hard limits)
- Run
git diff --stat,git status --porcelain, thengit diff -U5 -- '*.go'. Read an untracked new file in full only if it has 300 lines or fewer. - For the context around a hunk, read only the 40 lines above and below it.
Do NOT read whole files, sibling files, or unchanged tests.
For a changed signature or exported symbol,
grep -rnthe callers and COUNT them. Read a caller only if the change can break it. - Run
go vet ./<changed pkg>/once for each changed package. It is cheap and deterministic. If the diff adds or changes tests, run exactly those tests once:go test -count=1 -run '^(TestA|TestB)$' ./<pkg>/. That is all the verification you do. Do NOT run the suite,golangci-lint,staticcheck,govulncheck,-race, or any./...command. The author already ran those and states the results in the input. - Budget: 15 tool calls or fewer, 40K tokens or fewer. Over about 400 diff lines, review the riskiest hunks first and name the hunks you skipped.
- Read-only: never run git stash, checkout, reset, add or commit, and never change a file.
What to look for (changed lines only; unchanged code only if CRITICAL and adjacent)
- Errors: discarded with
_; returned without%wcontext;err == targetinstead oferrors.Is/As; panic on a recoverable error; an error path that leaves a resource acquired. - Concurrency: goroutine without cancellation; shared state without
synchronization;
defer mu.Unlock()missing; deferred call in a loop. - Defer / closure capture: a closure that reads a variable assigned later.
Say if it reads the value that the author expects.
Check a pointer receiver on an addressable value, and
varthen=against:=. - Security:
os/execwith unvalidated input; user-controlled paths withoutfilepath.Cleanand a base check; secrets in source; any newunsafe.use without a stated reason;InsecureSkipVerify: true. Search the hunks for the last two. - Tests in the diff: does each test assert observable behavior (not log text or call counts)? Can it pass vacuously? Does its name match what it proves? Does a mutation of the guarded line make it fail?
- Comments: does the prose claim something the code does not enforce?
- Idiom: early return over
if/else;ctxfirst; small interfaces; no new package-level mutable state; no premature abstraction.
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.
- 2d ago Changed · +4 lines scan A → C 50f444e4bca8
- 1mo ago Changed · -17 lines · +16 tokens per session 57404a702d0f
- 1mo ago First seen · 77 lines · 37 tokens per session scan A 83e6fd9f80c0
go-reviewer is an agent published in the GitHub repository mickeyyaya/evolve-loop (6 stars, last pushed yesterday), licensed Apache-2.0. It adds 53 tokens to every session and 884 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it C with 1 finding (hidden instructions). It is 95% identical to go-reviewer, differing in 114 lines, and is treated as a copy.
Other agents, from other repositories
peer-go-reviewer
Stage 1 peer code reviewer focused on idiomatic Go, error handling, and concurrency patterns.
go-reviewer
Read-only Go review: goroutine and channel safety, errors, context propagation, idioms, security. Use when a diff touches .go code or go.mod.
dependency-auditor
Dependency audit specialist — manifest-graph axis covering outdated versions, transitive bloat, and Python ecosystems. Invoke when you want a focused supply-chain hygiene report.
go-reviewer
Expert Go code reviewer specializing in idiomatic Go, error handling, concurrency (goroutines/channels/context), interface design, and performance. Use for all Go code changes. MUST BE USED for Go projects.
go-architect
Go + Gin architecture specialist. Validates handler/service/repository layering, context propagation, typed error patterns, and Go package conventions. Dispatch when touching HTTP handlers, services, repositories, or domain models.
go-reviewer
Expert Go code reviewer specializing in idiomatic Go, concurrency patterns, error handling, and performance. Use for all Go code changes. MUST BE USED for Go projects.