Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Manage crisis communication across prevention, response, and recovery phases using SCCT theory and crisis statement frameworks. Use this skill when the user faces a PR crisis, needs to draft a crisis response, plan crisis preparedness, or manage negative public attention — even if they say 'we're getting bad press', 'how do we respond to this scandal', 'draft a public statement', or 'prepare for potential backlash'.
.claude/skills/asgard-ai-platform-pr-crisis-communication/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 37% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 80% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 31% | 0% |
IRON LAW: Respond in the Golden Hour
The first public response must come within 1 hour of the crisis becoming
public. Silence creates a vacuum that others fill with speculation.
A holding statement ("We are aware and investigating") is better than
no statement. Perfection is the enemy of timeliness.1. Prevention (before crisis)
2. Response (during crisis)
3. Recovery (after crisis)
| Crisis Type | Attribution of Blame | Response Strategy | |------------|---------------------|-------------------| | Victim (natural disaster, rumor) | Low — organization is also a victim | Deny / Diminish | | Accidental (technical error, product defect) | Medium — unintentional | Diminish / Rebuild | | Preventable (human error, organizational failure) | High — could have been avoided | Rebuild (full apology + corrective action) |
[Concern] We are deeply concerned about [specific situation] and our
thoughts are with [affected parties].
[Facts] Here is what we know: [factual summary, no speculation].
[Commitment] We are taking the following immediate actions:
1. [Action 1]
2. [Action 2]
[Control] We have activated [response team/process] and will provide
updates [frequency and channel].
[Contact] For questions, contact [spokesperson, channel].markdown# Crisis Response Plan: {Situation} ## Crisis Assessment - Type: Victim / Accidental / Preventable - Severity: Low / Medium / High - Stakeholders affected: {list} - Media attention level: {current state} ## Immediate Response (< 1 hour) - Holding statement: {draft} - Spokesperson: {who} - Internal notification: {who needs to know} ## Full Response (< 24 hours) - Full statement: {draft using 3C framework} - Key messages (3 max): ... - Channel strategy: {where to publish} - Q&A preparation: {anticipated questions + answers} ## Recovery Plan - Corrective actions: ... - Trust-rebuilding steps: ... - Timeline: ...
references/social-crisis.md| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 23,397 | 18,753 | -20% | 1 | 1 | 0% | 3,671 | 3,792 | +3% | 0 | 0 | — |
case-02 | fail→pass | 20,928 | 19,723 | -6% | 1 | 1 | 0% | 3,547 | 3,905 | +10% | 0 | 0 | — |
case-03 | fail→fail | 52,917 | 26,591 | -50% | 1 | 1 | 0% | 8,261 | 5,017 | -39% | 0 | 0 | — |
case-04 | fail→pass | 19,780 | 18,795 | -5% | 1 | 1 | 0% | 2,746 | 3,758 | +37% | 0 | 0 | — |
case-05 | fail→pass | 11,594 | 18,951 | +63% | 1 | 1 | 0% | 1,938 | 3,485 | +80% | 0 | 0 | — |
case-06 | fail→pass | 15,790 | 13,986 | -11% | 1 | 1 | 0% | 2,399 | 3,151 | +31% | 0 | 0 | — |
case-07 | pass→pass | 12,280 | 17,260 | +41% | 1 | 1 | 0% | 1,789 | 3,345 | +87% | 0 | 0 | — |
case-08 | pass→pass | 12,990 | 12,257 | -6% | 1 | 1 | 0% | 2,062 | 2,827 | +37% | 0 | 0 | — |
case-09 | pass→pass | 15,522 | 19,224 | +24% | 1 | 1 | 0% | 2,247 | 3,657 | +63% | 0 | 0 | — |
case-10 | pass→pass | 14,351 | 16,782 | +17% | 1 | 1 | 0% | 2,216 | 3,479 | +57% | 0 | 0 | — |
case-11 | pass→pass | 14,039 | 19,664 | +40% | 1 | 1 | 0% | 2,226 | 3,920 | +76% | 0 | 0 | — |
case-12 | fail→pass | 13,311 | 17,120 | +29% | 1 | 1 | 0% | 2,014 | 3,506 | +74% | 0 | 0 | — |
case-13 | pass→pass | 13,016 | 19,196 | +47% | 1 | 1 | 0% | 2,063 | 4,037 | +96% | 0 | 0 | — |
case-14 | pass→pass | 12,043 | 17,254 | +43% | 1 | 1 | 0% | 1,902 | 3,238 | +70% | 0 | 0 | — |
case-15 | pass→pass | 12,124 | 15,205 | +25% | 1 | 1 | 0% | 1,916 | 3,166 | +65% | 0 | 0 | — |
case-16 | pass→pass | 12,792 | 18,299 | +43% | 1 | 1 | 0% | 2,271 | 3,583 | +58% | 0 | 0 | — |
case-17 | pass→pass | 14,099 | 10,905 | -23% | 1 | 1 | 0% | 2,079 | 2,649 | +27% | 0 | 0 | — |
case-18 | pass→pass | 11,076 | 13,312 | +20% | 1 | 1 | 0% | 1,724 | 2,926 | +70% | 0 | 0 | — |
case-19 | pass→pass | 13,353 | 8,949 | -33% | 1 | 1 | 0% | 1,891 | 2,225 | +18% | 0 | 0 | — |
case-20 | pass→pass | 8,470 | 11,050 | +30% | 1 | 1 | 0% | 1,429 | 2,455 | +72% | 0 | 0 | — |
case-21 | pass→pass | 14,988 | 12,741 | -15% | 1 | 1 | 0% | 2,378 | 2,713 | +14% | 0 | 0 | — |
case-22 | pass→pass | 19,727 | 18,696 | -5% | 1 | 1 | 0% | 2,682 | 3,881 | +45% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +27 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.