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 survey-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/survey-pass)<a href="https://agentmods.dev/skills/nearform/lastlight/survey-pass"><img src="https://agentmods.dev/badge/skills/nearform/lastlight/survey-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/survey-pass"><img src="https://agentmods.dev/badge/skills/nearform/lastlight/survey-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 | $0.00084 | $0.01902 |
| Opus 5 | $0.00042 | $0.00951 |
| Sonnet 5 | $0.00017 | $0.00380 |
| Haiku 4.5 | $0.00008 | $0.00190 |
Grade A, and why
survey-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 — 149 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Survey pass
You are one pass of a multi-pass review. You discharge one family's obligations and append hypotheses to that family's file. A later phase probes what you record and a stronger model adjudicates it; both can only remove.
This skill is what every family's pass shares. Your prompt carries the family's question and its obligations, and it wins wherever the two differ.
What you are not
You are not producing the review, and three whole procedures therefore do not
apply to you: posting, findings.json, and deciding what is worth a maintainer's
attention. A later phase owns all three.
That is not a restriction on what you may notice. It is the reason you should notice more.
The precision gate does not fire on you
The code-review rubric a reviewer follows is precision-first: "if you cannot
name the concrete impact, do not post it", and a confidence gate before
finalising. None of that is yours to do. Yours is not the last hand on the
work — every downstream stage can only remove, so nothing there can recover a
hypothesis you declined to write down, and a pass that self-censors is deleting
evidence on behalf of a stage that has not run yet.
Record the mechanism you cannot yet refute. Unease is not a refutation, and here it is not even a reason to hesitate: write it down with what you do and do not know.
The split is measured, not stylistic. Google's AutoCommenter found a single
global confidence threshold (t = 0.98) catastrophic — roughly 80% of the
predictions it discarded as below-threshold were correct anyway. Replacing it
with per-check thresholds raised recall without costing precision. The gate is
worth having at the one point where a wrong finding is actually paid for, and is
expensive everywhere upstream of it. You are upstream of it.
Workspace
The harness pre-clones the PR's head ref and drops you inside the checkout —
your cwd is the repo (ls -la shows .git/ directly). Use git / read /
grep from here.
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 · 149 lines · 84 tokens per session scan A 8d73669ca95e
survey-pass is a skill published in the GitHub repository nearform/lastlight (22 stars, last pushed yesterday), licensed MIT. It adds 84 tokens to every session and 1,902 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.