Social Crisis Response Planner
Build a severity-based social crisis playbook with queue pauses, approved statements, approvals, and recovery steps.
The skill explicitly limits pausing, publishing, and replying to humans, treats pasted material as untrusted, and uses proposal events, sole-writer rules, and reconciliation to reduce unauthorized or irreversible actions. However, user confirmation for pausing scheduled and paid activity, minimization and retention of sensitive crisis material, and rollback authorization are not fully explicit, so 7 points are deducted.
The contract, inputs, outputs, routing rules, missing-baseline behavior, Estimated labels, and stand-down flow are fairly detailed. Reliability is reduced because the skill depends on unprovided shared skills, memory structures, and runtime protocols, with no skill-specific tests or concrete failure messages for abnormal inputs; static calibration therefore keeps the score conservative.
Triggers, scenarios, exclusions, inputs, outputs, and launch-window/deliverability routing are clearly stated, with Chinese-platform examples. Boundaries for platform coverage, organizational variation, legal review, and regional differences are not well evidenced; mainland-China use appears to depend on user-provided exports rather than fully reachable native integrations, costing 5 points.
The skill has structured frontmatter and layered Quick Start, Skill Contract, Data Sources, Instructions, Next Best Skill, and termination sections, and states Apache-2.0, version, and host compatibility. It lacks a skill-specific changelog, named maintenance owner, or explicit update path, while some assumptions are delegated to external shared files, costing 5 points.
The claimed output covers the severity ladder, pause rule, holding-statement library, approval matrix, stand-down criteria, and retro template, which could produce a usable protocol. The actual library and templates are specified rather than supplied, and the workflow requires user data and human execution; no directly verified representative output is present, so 9 points are deducted.
The files provide auditable input labels, event paths, cross-skill references, and repository-level CI/test material. There is no skill-specific key-path test coverage, third-party corroboration, or execution evidence, so only limited static verifiability is justified and 6 points are deducted.
- Pausing all scheduled content and paid amplification is a high-impact external action; add explicit human confirmation, authorization ownership, and per-channel rollback records before production use.
- Rules such as journalist or regulator contact meaning at least severity 3 and the 1.5x/24-hour stand-down threshold are Estimated defaults, not substitutes for legal, PR, or platform-policy judgment.
- Screenshots, DMs, and forwarded emails may contain personal or confidential data; define redaction, access control, and retention limits before saving them to memory.
- This review did not execute scripts, verify referenced files, or verify the publisher's identity.
What does this skill do, and when should you use it?
Crisis Response Planner is a host-phase social skill in the Aaron Marketing Skills collection. It creates or activates a social crisis protocol with a 1–5 severity ladder tied to mention velocity, sentiment changes, journalist or regulator contact, and employee-conduct or safety incidents. The playbook requires pausing all scheduled posts and paid amplification, then defines holding statements, spokesperson and approval roles, do-not-post rules, all-clear criteria, and a post-crisis retrospective. It produces planning and incident-runbook artifacts; it does not automate posting, replies, or direct messages.
Reads the 7-day listening baseline, channel dossiers and states, scheduled social content, paid-amplification calendars, launch-window dates, and user-provided incident evidence; distinguishes protocol drafting, live-incident triage, and stand-down modes; builds five severity levels with thresholds, owners, first actions, statement classes, and update cadences; defines the first mechanical action as pausing all scheduled posts and paid amplification; creates pre-approved holding statements for product failure, employee conduct, account compromise, and brand misinformation; writes do-not-post rules, spokesperson and approval matrices, all-clear criteria, and a retrospective template; requires a social-quality-auditor pre-publish rerun before unpausing and submits dated pause or unpause marker proposals for channel reconciliation.
- A social lead needs a prewritten crisis protocol for channels such as LinkedIn, X, or Xiaohongshu.
- Mentions reach several times the seven-day baseline and a journalist contacts the team, requiring severity and first-action guidance.
- A marketing team must pause scheduled content and paid amplification while assigning drafting, approval, posting, and press-speaking responsibilities.
- An incident has ended and the team needs to reconcile pause markers, recheck queued content, and define the return to normal posting.
- A company wants its everyday inbox escalation taxonomy to end in a named crisis path rather than an unspecified escalation.
What are this skill's strengths and limitations?
- Combines severity, queue pausing, statement approval, update cadence, and recovery criteria in one operational playbook.
- Keeps crisis actions human-executed, with no posting, reply, or DM automation in the social discipline.
- Labels missing baselines, user-provided evidence, and estimated thresholds instead of presenting unsupported certainty.
- Defines handoffs to social-pulse-monitor, social-quality-auditor, channel-registry, and launch-day-conductor.
- It does not directly pause schedulers, change ad-platform state, or publish replies; humans must perform those actions.
- Default velocity thresholds are Estimated and require tuning; without a seven-day baseline, the skill cannot invent one.
- A complete workflow depends on access to project memory, channel state, queues, and the repository's shared scripts and protocols.
- The source provides no independent test suite, platform coverage results, or outcome data for this individual skill.
How do you install this skill?
For an Agent Skills-compatible host, install the repository with: npx skills add aaron-he-zhu/aaron-marketing-skills. To retrieve a single skill on a generic host, the repository documents: npx skills add aaron-he-zhu/aaron-marketing-skills -s crisis-response-planner. The collection also documents Claude Code plugin installation and git clone; the source does not specify a separate placement procedure for this skill.
How do you use this skill?
Give the compatible host a concrete request, such as: "Draft our social crisis protocol: channels LinkedIn + X + 小红书, team of 2, spokesperson = founder, baseline from last week's pulse sweep." For live triage, provide the mention multiple relative to the seven-day baseline and any journalist or regulator contact. For recovery, request: "Run the stand-down: reconcile the pause markers, re-run the pre-publish gate on the queued posts, then un-pause." During an active launch window, route incident handling to launch-day-conductor; route email deliverability incidents to deliverability-qa.
How does this skill compare with similar options?
Compared with deliverability-qa, this skill handles social crises rather than email blocklists or spam-rate spikes. Compared with launch-day-conductor, it stands down during an active launch window because that skill owns launch-day incident handling. Compared with social-quality-auditor, it writes the crisis protocol and requests the pre-publish rerun but does not score the ECHO profile or run vetoes.