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 sergei-aronsen/claude-code-toolkit --skill reality-checkgit clone --depth 1 https://github.com/sergei-aronsen/claude-code-toolkitWrote 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/sergei-aronsen/claude-code-toolkit/reality-check)<a href="https://agentmods.dev/skills/sergei-aronsen/claude-code-toolkit/reality-check"><img src="https://agentmods.dev/badge/skills/sergei-aronsen/claude-code-toolkit/reality-check/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/sergei-aronsen/claude-code-toolkit/reality-check"><img src="https://agentmods.dev/badge/skills/sergei-aronsen/claude-code-toolkit/reality-check.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 | $0.00042 | $0.00918 |
| Opus 5.5 | $0.00017 | $0.00367 |
| Sonnet 5.5 | $0.00008 | $0.00184 |
| Haiku 4.5 | $0.00004 | $0.00092 |
Grade A, and why
reality-check 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 1mo 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 — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reality Check (Post-Ship)
Iron law
NO SHIP IS COMPLETE WITHOUT RUNTIME VERIFICATION AGAINST PRODUCTION.
Code merged + CI green + deploy succeeded ≠ feature works for users. Reality check is the only thing that proves it.
When this skill activates
- Immediately after
/gsd-shipcompletes - When user says: "did it work", "verify deploy", "check prod"
- 24 hours after ship for delayed regression check
- After ANY hotfix to production
The procedure
Step 1: Smoke test (within 1 minute of deploy)
Run Playwright e2e against deployed URL, NOT localhost:
PROD_URL=https://yourapp.com npm run test:e2e:prod
Expected: green. If red — rollback immediately, investigate next.
Step 2: Error rate (within 1 hour)
Check Sentry dashboard for last 1 hour vs last 24 hours baseline:
Last 1h errors: N
Last 24h average / hour: M
If N > 2*M → red flag, investigate
If N > 5*M → rollback now
Step 3: User behavior (within 6 hours)
Check Posthog (or similar) for conversion rate change on critical funnel:
Pre-ship baseline (avg of last 7 days): 12.3% signup→first-action
Post-ship 6h window: 11.8% (within ±1%) → OK
Post-ship 6h window: 8.1% → red flag
Post-ship 6h window: <5% → rollback
Step 4: 24-hour delayed regression
Some bugs surface after timezone rollover, weekly cron jobs, or batch processes. Check at 24h post-ship:
- Sentry error count for last 24h (full day cycle)
- Customer support tickets count
- Posthog funnel for full day
Tooling
- Playwright — pre-ship smoke + post-ship runtime smoke
- Sentry — error monitoring (set up SLO alerts: page on >5× baseline)
- Posthog — funnel monitoring (alert on >20% conversion drop)
- Plausible — alternative to Posthog (privacy-first, simpler)
If any of these are NOT installed — see skills/production-observability/SKILL.md first. Reality check requires observability infrastructure.
What to do when red
Step 1 (Playwright fail) → Rollback. Investigate logs. Hotfix or revert.
Step 2 (Sentry spike) → Triage top error. If user-facing: rollback. If internal: hotfix.
Step 3 (conversion drop) → Don't rollback yet. Check session replay. Often UX bug, not crash.
Step 4 (delayed) → Same triage as Step 2/3 but with more context.
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.
- 1mo ago First seen · 101 lines · 42 tokens per session scan A f0b3769cfb23
reality-check is a skill published in the GitHub repository sergei-aronsen/claude-code-toolkit (5 stars, last pushed 2mo ago), licensed MIT. It adds 42 tokens to every session and 918 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
verify-omnigent
Drive Omnigent the way a user does and prove a behavior with recorded evidence, using an isolated server, runner, and mock model plus a feature map of every user entry point. Load before reproducing a user-facing bug, before claiming a fix works, or when reviewing whether a change covered every surface (session vs.…
test-master
Generates test files, creates mocking strategies, analyzes code coverage, designs test architectures, and produces test plans and defect reports across functional, performance, and security testing disciplines. Use when writing unit tests, integration tests, or E2E tests; creating test strategies or automation…
eval-dataset-design
A guide to creating reliable test sets for AI agents. Each evaluation task includes the request, the simulated user's behaviour, a reset starting state, and a success check that can be independently verified.
dogfood
Systematically explore and test a mobile app on iOS/Android with agent-device to find bugs, UX issues, and other problems. Use when asked to dogfood, QA, exploratory test, find issues, bug hunt, or test this app on mobile.
cua-driver
Use Cua Driver for desktop or browser tasks that are awkward or unavailable through Bash/APIs, or when the user explicitly wants GUI interaction: app testing, visual bug reproduction, form filling, calendar entry, screenshots, and demo recording. Also covers Cua setup; not OpenAI Codex Computer Use or web research.
critique
Adversarial design critique of the current work. Render it, look at it, and argue for rejection. Run after the gates are green, never instead of them.