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 minhnv0807/ai-business-skills --skill 66-crisis-playbook-globalgit clone --depth 1 https://github.com/minhnv0807/ai-business-skillsWrote 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/minhnv0807/ai-business-skills/66-crisis-playbook-global)<a href="https://agentmods.dev/skills/minhnv0807/ai-business-skills/66-crisis-playbook-global"><img src="https://agentmods.dev/badge/skills/minhnv0807/ai-business-skills/66-crisis-playbook-global/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/minhnv0807/ai-business-skills/66-crisis-playbook-global"><img src="https://agentmods.dev/badge/skills/minhnv0807/ai-business-skills/66-crisis-playbook-global.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Anti-Refusal · line 28 Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.Fix: Remove instructions that suppress warnings, disclaimers, or ethical commentary. Let the agent surface safety-relevant caveats to the user.
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 | $0.00165 | $0.02518 |
| Opus 5 | $0.00082 | $0.01259 |
| Sonnet 5 | $0.00033 | $0.00504 |
| Haiku 4.5 | $0.00016 | $0.00252 |
Grade A, and why
66-crisis-playbook-global 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 — 193 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Crisis Playbook (Global)
A crisis arrives without warning, so the playbook must exist before it is needed. Three ground rules: (1) do not react before you have facts — most crises look worse than they are in the first 30 minutes; (2) do not stay silent past 4 hours on a spreading crisis, because silence reads as admission; (3) handled well, a crisis builds trust; handled badly, it destroys it permanently.
Information gathering
Read the 15-social-listening-global report if available (early detection). Before doing anything, answer 4 questions:
- What type of crisis is this? Match it against the L1-L5 table below.
- How fast is it spreading? Ten people or ten thousand? One channel, or has it jumped channels?
- Where did it start? An unhappy customer, a creator or influencer, a competitor, a regulator, or the press?
- What is actually true? Which fact is disputed, do you have evidence, and do you have enough to respond at all?
If you do not have enough information: publish one short line confirming you are investigating and commit to a specific update time. Do not go silent, and do not respond to unverified claims.
Principles
- Assess first, react second. Thirty to sixty minutes of investigation beats a wrong answer in five minutes.
- Never silent past 4 hours on a spreading crisis.
- Do not delete legitimate negative comments (clear spam and hate speech excepted) — deleting escalates.
- Do not argue in public. Move the complainant to a private channel.
- Do not respond while emotional. Draft it, wait 30 minutes, reread, then send.
- Do not promise what you cannot deliver, and do not minimize the facts.
- L3 and above must go to the leader. Never self-handled.
- Regulated categories escalate straight to legal. In health, finance, insurance, credit, and crypto, any challenge to a claim goes directly to legal review — do not draft a public response first.
Crisis classification
| Level | Type | Example | Priority | Response window |
|---|---|---|---|---|
| L1 | Operational — campaign or system failure, severe KPI miss | Ads not delivering, landing page down, tracking broken | High | 2-4 hours |
| L2 | Content — inaccurate or misleading content published | Misleading caption, wrong figure, inappropriate image | High | 2-4 hours |
| L3 | Reputation — customer complaint spreading | Viral one-star review, exposé thread | Critical | 1-2 hours |
| L4 | Brand attack — coordinated | Mass fake reviews, competitor smear, brigading | Critical | Same day |
| L5 | Legal / compliance / data | Deceptive-advertising or undisclosed-endorsement complaint, personal data breach, IP infringement, platform enforcement action | Critical + counsel | Immediate |
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 · 193 lines · 165 tokens per session scan A 9ebf8e40a1bf
66-crisis-playbook-global is a skill published in the GitHub repository minhnv0807/ai-business-skills (572 stars, last pushed 26d ago), licensed MIT. It adds 165 tokens to every session and 2,518 once invoked, about $0.0008 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 2026-09-03.
Other skills, from other repositories
kanban-video-orchestrator
Plan and run multi-agent video production pipelines.
sdlc-review
Review Kanban handoffs and route verified outcomes.
teams-meeting-pipeline
Teams meeting summaries, job replay, Graph subscriptions.
batch
Research and plan a large-scale change, then execute it in parallel across 5–30 isolated worktree agents that each open a PR.
meeting-action-items
Turn meeting notes into cited decisions, owners, tickets.
subagent-dag-orchestration
Use when a task breaks into several distinct steps that a separate sub-agent could each carry out. A DAG node dispatches to an agent on the roster and cannot call your own tools, so steps that are just your own tool calls (reading files, small edits, running commands) are not a DAG for being many - do those yourself.…