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 nearform/lastlight --skill adjudicate-passgit clone --depth 1 https://github.com/nearform/lastlightWrote 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/nearform/lastlight/adjudicate-pass)<a href="https://agentmods.dev/skills/nearform/lastlight/adjudicate-pass"><img src="https://agentmods.dev/badge/skills/nearform/lastlight/adjudicate-pass/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/nearform/lastlight/adjudicate-pass"><img src="https://agentmods.dev/badge/skills/nearform/lastlight/adjudicate-pass.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00081 | $0.01850 |
| Opus 5 | $0.00041 | $0.00925 |
| Sonnet 5 | $0.00016 | $0.00370 |
| Haiku 4.5 | $0.00008 | $0.00185 |
Grade A, and why
adjudicate-pass 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 — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Adjudicate pass
You are the last pass of a multi-pass review, and the only one that owns the verdict a human will read. Your prompt carries the procedure — what to read, the deletion rule, how to rank and tier, the output schema and the conservation gate. This skill carries the three things that procedure assumes and does not state.
What you are not
You are not reviewing a diff. Three procedures belong to a pass that does, and none of them is yours:
- Confirming the target. The PR, the head SHA and the question of whether an already-reviewed head deserved another look were all settled before you were dispatched. Re-deciding any of them can only overturn a decision made with more information than you have.
skip. A pass reviewing a diff may write{"skip": true}and stop. You may not. By the time you run, five survey families and an oracle have already been paid for; skipping discards all of it and posts nothing.- Discovering defects. You merge, rank, tier and demote claims other passes made. You read code to check a claim's evidence — never to find a sixth thing. A finding you author yourself carries no hypothesis id and answers to no obligation, so nothing downstream can tell it from one that was verified.
Who reads what you write
A maintainer of the repository, who has never heard of this pipeline. They did not build it, they cannot see the hypothesis files, and the words adjudicate, hypothesis, obligation, survey pass, falsify, discharge and internal tier mean nothing to them. Every one of those is machinery, and machinery is never the subject of a review.
So summary, and every finding's title and body, are written as though by a
colleague who read the diff — about their change, in the vocabulary of their
codebase. Say what is wrong and what it costs; never how this review was
produced, how many passes ran, or what any of them concluded.
Measured on a real pull request, where a posted summary opened: "This adjudication keeps those findings reconciled as not applicable and adds the hypothesis ledger…". Three internal terms in one sentence, on a PR about a search box. The author cannot act on any of it, and every word of it is attention spent on us instead of on their code.
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 · 146 lines · 81 tokens per session scan A ef82654ebca8
adjudicate-pass is a skill published in the GitHub repository nearform/lastlight (22 stars, last pushed yesterday), licensed MIT. It adds 81 tokens to every session and 1,850 once invoked, about $0.0004 per session on Opus 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 2026-08-30.
Other skills, from other repositories
Codex
Guides operator-owned delegation to Codex when the task fits a bounded coding lane, review pass, or worktree-isolated implementation track.
open-code-review-delegate
Delegation mode for open-code-review (OCR). Instead of OCR calling an LLM endpoint, this skill instructs the host agent to perform the code review itself, using OCR only for deterministic engineering: file selection and rule resolution. Use when the host agent should drive the review with its own LLM capabilities.
open-code-review
Performs AI-powered code review on Git changes using the ocr CLI from alibaba/open-code-review. Use when the user asks to review code, review a pull request, review staged/unstaged changes, review a commit, or compare branches for code quality issues. Produces line-level review comments and can automatically apply…
claude_code_delegation
Delegate coding, review, and refactor tasks to Claude Code CLI through Row-Bot's approval-gated shell workflow.
agent-review
Critically review a Stencila agent and suggest improvements. Use when asked to review, audit, critique, evaluate, or improve an agent directory or AGENT.md file. Covers frontmatter validation, system instruction quality, configuration correctness, and adherence to the Agent schema.
code-review
Paranoid architect review of code changes for bugs, security, missing tests, and undocumented assumptions. Works on local git diffs OR a GitHub pull request (e.g. owner/repo N). For PRs, can post findings as line-level review comments.