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 jonZlotnik/how-to-agent --skill crash-earlygit clone --depth 1 https://github.com/jonZlotnik/how-to-agentWrote 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/jonzlotnik/how-to-agent/crash-early)<a href="https://agentmods.dev/skills/jonzlotnik/how-to-agent/crash-early"><img src="https://agentmods.dev/badge/skills/jonzlotnik/how-to-agent/crash-early/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/jonzlotnik/how-to-agent/crash-early"><img src="https://agentmods.dev/badge/skills/jonzlotnik/how-to-agent/crash-early.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 | $0.00050 | $0.00691 |
| Opus 5 | $0.00025 | $0.00345 |
| Sonnet 5 | $0.00010 | $0.00138 |
| Haiku 4.5 | $0.00005 | $0.00069 |
Grade A, and why
crash-early 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.
How it starts
The opening of the file, as written. The whole thing — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Crash Early
A program that has wandered into an impossible state is no longer your program. Every line it executes afterwards is operating on a lie. The earlier and louder it stops, the smaller the blast radius, the more honest the diagnostic, and the less time you spend later wondering how data got into that shape.
When to use
- Writing any
try/catch,except,Result::unwrap_or,?? defaultValuebranch - Adding a "just in case"
if value is Noneguard - A library returns an unexpected error and you're tempted to log-and-continue
- Reviewing code that catches a broad exception class
Protocol
Classify every error path into one of three buckets, then act accordingly:
- Bug ("impossible" state). Assertion or hard abort. This is a contract violation ([[design-by-contract]]) — your program now lies if it keeps running. Examples: invariant broken, null where the type forbids it, "this branch is unreachable".
- Exceptional (rare, recoverable, real). Raise/return an error type. The caller must explicitly handle or propagate. Examples: network timeout, disk full, lock contention.
- Expected (normal control flow). Handle locally with a normal return value. Examples: user not found in a lookup, end-of-stream, optional config missing.
Confusion among the three is the source of most "silent corruption" bugs. The cure is to name the bucket in the comment, type, or error name. BugError, ConfigMissing, Result<T, NetworkError>.
Specific anti-patterns to refuse
except Exception: pass— bug-class swallowing. Crash.if x is None: return Nonefor an x that should never be None — invents a sentinel; future code can't tell bug from expected.- A retry loop around a deterministic error — retrying a bug just makes it slower.
- A fallback that masks the failure ("oh well, we'll use the default") when the failure means data loss.
Red flags (rationalizations to reject)
- "We can't let it crash in production." — Crashing on a violated contract is better than continuing on corrupted state. Recover at the process boundary (supervisor, retry-the-request), not by lying to your own code.
- "I'll log and continue." — Logging is not handling. Now you have corrupted state and noise.
- "Defensive programming is safer." — Defending against impossible states creates the conditions for them to go unnoticed.
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 · 46 lines · 50 tokens per session scan A ebbfbe2b8a10
crash-early is a skill published in the GitHub repository jonZlotnik/how-to-agent (2 stars, last pushed 1mo ago), licensed MIT. It adds 50 tokens to every session and 691 once invoked, about $0.0003 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 2026-08-31.
Other skills, from other repositories
debug-optimize-lcp
Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…
systematic-debugging
Use when debugging a failing test, build error, or runtime issue that isn't immediately obvious. Guides a 4-phase root cause analysis instead of random fix attempts.
diagnose
Trace from a reproduced symptom to the source code that causes it. Pin the specific file and approximate line, rate confidence in the cause and clarity of the fix independently, and always propose a concrete fix.
repro-admin
Reproduce an EmDash admin UI bug. Attach a container, start the demo dev server, drive the admin with agent-browser using the dev-bypass session, and capture the reproduction as screenshots plus a replayable transcript.
log-error-digest
Analyze log files to troubleshoot errors, identify peak error periods, and produce error clustering, frequency statistics, and time distribution reports. Supports JSON, syslog, and Nginx formats with automatic detection. Use when a user uploads a .log file and asks to analyze errors, find patterns, debug issues, or…
byted-util-volcengine-detect-retry
An orchestration workflow for Volcengine Cloud Detect, a service that checks websites or network endpoints from test locations.