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/harnessworks/harness-starter-kitWrote 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/commands/harnessworks/harness-starter-kit/harness-review)<a href="https://agentmods.dev/commands/harnessworks/harness-starter-kit/harness-review"><img src="https://agentmods.dev/badge/commands/harnessworks/harness-starter-kit/harness-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/commands/harnessworks/harness-starter-kit/harness-review"><img src="https://agentmods.dev/badge/commands/harnessworks/harness-starter-kit/harness-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 | $0.00000 | $0.02542 |
| Opus 5 | $0.00000 | $0.01271 |
| Sonnet 5 | $0.00000 | $0.00508 |
| Haiku 4.5 | $0.00000 | $0.00254 |
Grade A, and why
harness-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 — 238 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/harness review
Review the current change set from an opposing harness-engineering perspective.
Harness Review is diagnostic by default. Do not create, edit, delete, move, format, or stage files while running this command unless the user explicitly asks you to apply fixes after seeing the review.
Run this review before commit or push when practical. If it runs after commit or push, record that timing so maintainers know whether it acted as a safety gate or as a post-hoc diagnostic.
Goal
Challenge whether the current work preserves the target repository as the source of truth, avoids unnecessary automation, keeps templates conservative, adds enforceable checks only where practical, updates durable memory when needed, and runs the right validation before completion.
Every review must check whether the current environment exposes a multi-agent or subagent review tool, record the reviewer mode, and report any fallback reason. The parent or orchestrator agent that runs this command owns the reviewer-mode and fallback decision. Do not depend on any specific agent runtime.
Invocation Modes
/harness review
Run the normal diagnostic review. At the start, check whether multi-agent or subagent tools are available in the current environment. If a subagent tool is available and permitted by the active runtime and tool instructions for this invocation, invoke a read-only reviewer subagent before writing the final review.
If the active tool instructions require an explicit user request to spawn a
subagent and the user only asked for /harness review, treat the tool as
present but not permitted, continue with a single-agent reviewer perspective,
and report the fallback reason.
/harness review sub-agent
Run the same diagnostic review, but treat this invocation as the user's explicit request to use a read-only reviewer subagent. If a multi-agent or subagent tool is available and permitted by the active runtime and tool instructions, invoke it before writing the final review.
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 · 238 lines · 0 tokens per session scan A 1bd1b9438eed
harness-review is a command published in the GitHub repository harnessworks/harness-starter-kit (113 stars, last pushed 2mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 2,542 tokens. 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 commands, from other repositories
context
Triage card for files, modules, or symbols — layer, hotspot, fix history, freshness (relationships, not source bytes).
doctor
Diagnose the Repowise setup — install, API keys, index/store drift — and optionally repair it.
investigate
Command → Agent → Skill orchestration for investigating and fixing bugs.
code-quality
Code review and bug prediction — find quality issues, style violations, and likely bugs.
review-hard
Review the current task changes for real correctness, safety, and maintainability risks before slice closure or PR prep, and recommend the smallest safe next step. Surfaces meaningful issues (not cosmetic feedback); not a replacement for external review systems. Returns no-op when the review would not materially…
verify
This page documents the livewiki verify CLI subcommand, which checks a generated wiki for the kinds of corruption that would break readers or CI pipelines.