A workflow for launching and supervising a Nightwatch specification run, including preflight checks, live monitoring, pausing, and skipping selected specifications.

It is for starting Nightwatch, watching its logs and state, pausing or stopping a run, and steering individual specifications.

It helps detect blocked runs, unsafe repository state, and problems that require intervention instead of silently letting a run continue.

watch — Skill for Claude Code by jasonm4130

Install

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.

Any agent

npx skills add jasonm4130/claude-skills --skill watch

Clone the repo

git clone --depth 1 https://github.com/jasonm4130/claude-skills

Made for: Claude Code.

Or install nightshift, the plugin that ships this one along with the rest of its 5 skills, 2 agents.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/jasonm4130/claude-skills/watch/github.svg)](https://agentmods.dev/skills/jasonm4130/claude-skills/watch)

Your own site

<a href="https://agentmods.dev/skills/jasonm4130/claude-skills/watch"><img src="https://agentmods.dev/badge/skills/jasonm4130/claude-skills/watch/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/jasonm4130/claude-skills/watch"><img src="https://agentmods.dev/badge/skills/jasonm4130/claude-skills/watch.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 147 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,772 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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	$0.00147	$0.01772
Opus 5	$0.00073	$0.00886
Sonnet 5	$0.00029	$0.00354
Haiku 4.5	$0.00015	$0.00177

Measured 5d ago against content hash a47a893de416, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

watch 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.

plugins/nightshift/skills/watch/SKILL.md · 117 lines

How it starts

The opening of the file, as written. The whole thing — 117 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Watch a Nightwatch run

The first night failed five ways in the harness and once in the code: a print-mode timeout that fabricated a result, an env var that silently forced default permission mode, an untrusted clone that dropped project settings, a tee that polluted the state channel, and a dirty-tree false positive on a file the acceptance command itself wrote — against one real spec defect (an Earnings-21 reference file with empty timestamp columns). All five harness failures are fixed in the engine now; this playbook is for the failures a fixed engine still can't prevent — a blocked branch, a spec that needs an edit mid-run, a night that needs to stop early — and for reading the state a running launcher writes so nothing has to be guessed.

Announce: "Using nightwatch:watch to launch and steer the run."

1. Preflight

Before starting run.sh, confirm all of:

The clone is clean (git status --porcelain empty) and trusted (claude -p in it reads project settings — an untrusted workspace silently drops them).
gh variable get LANDING_STATE (from inside the clone; the launcher's default STATE_VAR) prints exactly run.
node <plugin>/nightwatch/lint-spec.mjs --specs-dir <specs-dir> --check <check> prints SPEC OK (<n> specs) for the whole queue — a spec defect caught here is free; caught at 2 a.m. it costs a unit.
1Password is unlocked if any command in the queue needs a secret.
$HOME/.local/state/nightwatch/<name>/launcher.lock is absent, or holds a pid that is no longer alive (kill -0 <pid> fails) — a live lock means another launcher already owns this clone. node <plugin>/nightwatch/init.mjs --report re-checks a setup already done — a clean way to confirm nothing has drifted (trust, the check command, the switch) before a launch.

2. Launch

caffeinate -i <plugin>/nightwatch/run.sh <name> [<specs-dir>]

<specs-dir> defaults to the config's SPECS, so a bare <name> is enough once init has run. In a herdr pane when HERDR_ENV is set, so the run has its own pane and this session never blocks on it; otherwise background it with stdin from /dev/null (the child reads stdin even given a prompt, and would otherwise eat this session's own input). --only <slug>[,<slug>...] narrows the queue for a resume or a single-spec relaunch.

Read the full file on GitHub · 117 lines

Changes

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 · 117 lines · 147 tokens per session scan A a47a893de416

Subscribe to this mod's changes

watch is a skill published in the GitHub repository jasonm4130/claude-skills (5 stars, last pushed 6d ago), licensed MIT. It adds 147 tokens to every session and 1,772 once invoked, about $0.0007 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 2026-09-06.

Testing ai-agents claude-code-plugin claude-skills plugin-marketplace

phx-work

Execute Elixir/Phoenix plan tasks with progress tracking. Use after phx-plan to implement features with mix compile and mix test verification after each step, or --continue to resume interrupted work.

oliver-kriska/claude-elixir-phoenix · 43 tokens

lab:autoresearch

Self-improving loop for plugin skills. Reads program.md, proposes one mutation per iteration, evaluates against deterministic scorer, keeps improvements via git, reverts failures. Targets weakest skill+dimension. Use with /loop for overnight runs.

oliver-kriska/claude-elixir-phoenix · 52 tokens

review-testing

Review test code for quality, design, and completeness after implementing a feature or fixing a bug. Use when the user asks to "review my tests", "check my test quality", "are these tests good enough", "review testing", or after completing a feature implementation that includes tests. Also use when tests feel brittle…

posit-dev/skills · 82 tokens

testing-r-packages

Best practices for writing R package tests using testthat version 3+. Use when writing, organizing, or improving tests for R packages. Covers test structure, expectations, fixtures, snapshots, mocking, and modern testthat 3 patterns including self-sufficient tests, proper cleanup with withr, and snapshot testing.

posit-dev/skills · 67 tokens

codex-loop

Fix Elixir/Phoenix code until Codex CLI review comes back clean — bounded review, fix, verify loop before opening a PR. Use when codex is installed and you want an external cross-model critic on your changes before pushing.