What is stop-or-continue?

A procedure for deciding whether an active work session should stop, continue, or record a handoff. It prioritizes correctness problems, such as a command or program changing while the session is running, over context-size concerns.

Use it when an engine, command, hook, or instruction file changes during a session, or when deciding whether an unfinished result is safe to hand off. It explains when restarting is required and when continuing is acceptable.

It prevents a session from continuing with stale code or rules and producing results that look like ordinary failures. It also ensures that work is documented before another session takes over.

How bredmond1019/claude-sdk-rs configures Claude Code and Codex

Reuse

Borrowing it

Nothing to install: this file belongs to bredmond1019/claude-sdk-rs. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/bredmond1019/claude-sdk-rs/main/.agents/skills/stop-or-continue/SKILL.md

Clone the repo

git clone --depth 1 https://github.com/bredmond1019/claude-sdk-rs

Made for: Claude Code, Codex.

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/bredmond1019/claude-sdk-rs/stop-or-continue/github.svg)](https://agentmods.dev/skills/bredmond1019/claude-sdk-rs/stop-or-continue)

Your own site

<a href="https://agentmods.dev/skills/bredmond1019/claude-sdk-rs/stop-or-continue"><img src="https://agentmods.dev/badge/skills/bredmond1019/claude-sdk-rs/stop-or-continue/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/bredmond1019/claude-sdk-rs/stop-or-continue"><img src="https://agentmods.dev/badge/skills/bredmond1019/claude-sdk-rs/stop-or-continue.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 110 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,480 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	$0.00110	$0.01480
Opus 5.5	$0.00044	$0.00592
Sonnet 5.5	$0.00022	$0.00296
Haiku 4.5	$0.00011	$0.00148

Measured 11d ago against content hash 1b67799b753f, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

stop-or-continue 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 11d 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.

.agents/skills/stop-or-continue/SKILL.md · 102 lines

How it starts

The opening of the file, as written. The whole thing — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Deciding to stop, continue, or write something down first

Paths below are relative to the brain root — the directory containing brain.toml, found by walking up from wherever you are. This skill is synced into every repo, so a repo-relative link would be wrong in most of them.

Three questions, in this order. Only the third one is about tokens, and most of the time you never reach it.

1. Is there a correctness reason to restart? — overrides everything

These are not efficiency calls. They hold at 40k context as firmly as at 400k, and continuing produces wrong results that look like ordinary failures, which is what makes them worth checking first.

Trigger	Why continuing is wrong
An engine, command or binary changed this session	Standing rule 10: the Workflow harness copies each engine `.js` at launch and runs that copy; `.claude/commands/.md` behave the same way. A fix on `main` does not change the running session. A stale engine emits the pre-fix command, the pre-fix failure returns, and it reads as an unreliable agent rather than a stale snapshot. Measured 2026-08-19: one block re-run four times* against an engine that never changed.
A locally-installed binary was rebuilt (`mev`, `bastion`)	Same shape one layer down, plus `emit-state --write` from a stale binary reverts generated boards to an older format. See `derive-state-safely`.
`settings.json`, hooks, or MCP config changed	The harness reads these at startup.
The operator changed a `CLAUDE.md` you already read	You are working from the superseded text. A mid-session diff notice is enough; a full rewrite is not.

If any of these fired, say so and name the trigger. Do not present it as a token decision — the operator should know the session is now producing untrustworthy evidence, not merely an expensive one.

2. Does the next chunk of work have a written entry point?

The gate is the artifact, not the number. If the next agent can start from a file, clearing is nearly free. If it cannot, the token count is the wrong thing to be looking at — write the artifact, then decide.

Read the full file on GitHub · 102 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.

11d ago First seen · 102 lines · 110 tokens per session scan A 1b67799b753f

Subscribe to this mod's changes

stop-or-continue is a skill published in the GitHub repository bredmond1019/claude-sdk-rs (23 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 110 tokens to every session and 1,480 once invoked, about $0.0004 per session on Opus 5.5. 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 2026-10-26.

Planning Productivity

stop-or-continue next to 6 skills from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
stop-or-continuebredmond1019/claude-sdk-rs	Skill	23	1mo ago	Apache-2.0	A	110	Claude Code, Codex
orbit-notionnexu-io/open-design	Skill	106k	today	Apache-2.0	A	117	Claude Code, Codex
agentmailpaperclipai/paperclip	Skill	111k	today	MIT	A	30	Claude Code, Codex
Cortexdanielmiessler/LifeOS	Skill	20k	yesterday	MIT	A	184	Claude Code
rowboat-spacesrowboatlabs/rowboat	Skill	18k	yesterday	Apache-2.0	A	60	Claude Code, Codex
feishuHmbown/CodeWhale	Skill	41k	5d ago	MIT	A	33	Claude Code, Codex
pinchtab-mcppinchtab/pinchtab	Skill	11k	10d ago	MIT	A	52	Claude Code, Codex

claude-sdk-rs: Skill for Claude Code