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/eddiebelaval/squireWrote 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/eddiebelaval/squire/finish)<a href="https://agentmods.dev/commands/eddiebelaval/squire/finish"><img src="https://agentmods.dev/badge/commands/eddiebelaval/squire/finish/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/eddiebelaval/squire/finish"><img src="https://agentmods.dev/badge/commands/eddiebelaval/squire/finish.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 | $0.01907 |
| Opus 5 | $0.00000 | $0.00954 |
| Sonnet 5 | $0.00000 | $0.00381 |
| Haiku 4.5 | $0.00000 | $0.00191 |
Grade A, and why
finish 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 5d 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 — 158 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/finish - run the whole quality chain
One command for Stage 8. Runs every gate in order, stops on a red, and reports each one's real state — including the ones that did not run.
This is the definition-of-done doctrine made executable: the last mechanical step is
yours. /finish is what "done" means for code.
The chain
1 /verify does it work? green baseline. everything below depends on it
2 /thrift is this the right way? change the APPROACH
3 /simplify what can go? shrink what survives
4 /polish is it finished? add the edges
5 /assay is it worth keeping? worth
6 /3pv is it what he asked? intent
Order is load-bearing: largest blast radius first. Approach → structure → edges → worth → intent. Each gate operates inside the result of the one before it. Reordered, the later gates spend their effort on code the earlier ones would have replaced.
Gates 2–4 CHANGE code. Gates 5–6 JUDGE it and change nothing.
Arguments
--from <gate>— start partway (/finish --from simplify). Baseline still runs.--only <gate>[,<gate>]— run just these, plus the baseline.--skip <gate>[,<gate>]— force-skip, recorded as SKIPPED(forced) in the report.--dry-run— every gate reports findings, nothing is applied.- Empty = all six on the current branch's diff. The common case.
THE ONE RULE
A gate that did not run must never render as a gate that passed.
Three terminal states, visually distinct, never collapsed:
| State | Means |
|---|---|
| PASS | ran to completion on the real target |
| SKIPPED | deliberately not run — always carries its reason |
| STOP | ran and failed, or could not run (UNVERIFIED) |
A run of six gates where four were skipped is a 2/6 run, and the summary line says so. It is not a green checkmark. This is the whole reason the runner exists as its own file rather than as a sentence telling you to run five commands: a human running them by hand skips silently and remembers a pass.
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.
- 5d ago First seen · 158 lines · 0 tokens per session scan A 0c0cb83565fe
finish is a command published in the GitHub repository eddiebelaval/squire (21 stars, last pushed 23d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,907 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 2026-09-03.
Other commands, from other repositories
audio-pass
Functional audio review for implementation, mix coverage, clarity, and player feedback.
ipd-tr
A command that runs a staged technical review process involving quality, testing, architecture, and requirements checks before a release decision. IPD-OMO is the process named in the add-on; its specialized role names are not explained further in the input.
wize-tea-review
Goal. A structured review of one story before its gate decision. Distinct from Shuri's wize-code-review (which audits code health). This one audits AC fulfillment, test discipline, and risk-spot coverage.
code-review
Comprehensive 3-pass code review of uncommitted changes. Covers quality (security, React, Node.js, performance), compliance (CLAUDE.md rules), and contract consistency (cross-repo API types).
verify-work
Validate built features through conversational UAT.
gx-e2e-review
You will be supplied with a commit or a directory or a test file. Your job is to review the E2E (Selenium/Playwright) tests.