菜单
Use when setting up an ongoing listening query, evidence log, cadence and escalation path. Produces social listening plan, query set and intelligence log; use `meta-sentiment-analysis` when that neighbouring contract is the closer match.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/peterbamuhigire/social-media-skills --skill meta-social-listening命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
基于 SOC 职业分类
正在显示 SKILL.md
| name | meta-social-listening |
| description | Use when setting up an ongoing listening query, evidence log, cadence and escalation path. Produces social listening plan, query set and intelligence log; use `meta-sentiment-analysis` when that neighbouring contract is the closer match. |
| metadata | {"portable":true,"compatible_with":["claude-code","codex"]} |
meta-sentiment-analysis is not the closer route before proceeding.meta-sentiment-analysis when its narrower output is requested.| Artefact | Source/provider | Required? | If absent |
|---|---|---|---|
| Brand terms, risks, competitors, languages and accessible sources | Client, approved systems, or dated platform exports | Yes | Stop the affected decision; request it or mark the field unknown and narrow the output. |
| Purpose, audience and approval boundary | Client brief or accountable owner | Yes | Return discovery questions; do not infer approval. |
| Artefact | Consumer | Acceptance condition |
|---|---|---|
| Social listening plan, query set and intelligence log | Client lead and next workflow owner | Every recommendation traces to an input, names an owner or next action, and marks assumptions and unassessed checks. |
| Evidence | Format | Acceptance condition |
|---|---|---|
| Decision and source register | Table in the deliverable | Each material claim records its source/date or is labelled unverified; missing evidence never becomes a pass. |
Read and search access to the supplied artefacts are required; calculation or file-rendering capability is optional. Planning and drafting are read-only with respect to client accounts and source records. Editing the deliverable requires explicit authorisation; publishing, production mutation, destructive action, spend, and certification claims require separate explicit authority and evidence.
If files, platform access, network, rendering, fonts, or calculation tools are unavailable, return the narrowest useful qualified social listening plan, query set and intelligence log. Mark each blocked check not assessed, state the consequence, and provide the exact evidence needed to resume. Never convert an unavailable check into a pass.
| Choice | Action | Failure or risk avoided |
|---|---|---|
| Brand terms, risks, competitors, languages and accessible sources is current and attributable | Produce the full social listening plan, query set and intelligence log and cite the evidence used. | Decisions based on stale or unrelated evidence. |
| A material input is missing or contradictory | Stop that decision, request clarification, or issue a labelled partial result. | Fabricated precision and false confidence. |
The requested outcome belongs to meta-sentiment-analysis | Route there and hand over the verified inputs already collected. | Neighbour collision and duplicated work. |
meta-sentiment-analysis if its contract is closer.The output is client-specific, uses British English and the stated market/currency, distinguishes observed fact from inference, exposes gaps, and gives a checkable acceptance condition. Recommendations must be feasible within the confirmed budget, capacity and permissions.
not assessed, its risk and the recovery input.meta-sentiment-analysis into this workflow. Fix: route the neighbouring output and hand over verified inputs.Given verified brand terms, the skill produces a social listening plan, query set and intelligence log with source dates and named assumptions. If that evidence cannot be accessed, it returns only the supported sections plus a recovery list; it does not fill gaps with East African defaults.
meta-sentiment-analysis for the neighbouring contract.anti-ai-slop during production.ai-slop-audit at the release checkpoint.Collect the following before generating the listening programme:
Social listening is the practice of monitoring online conversations about your brand, your competitors, and your industry in order to surface intelligence you would otherwise miss. It means searching systematically for mentions of your business name, your products, your key people, and your sector — across social media platforms, news sites, blogs, forums, and review sites — and then interpreting what you find to inform decisions. Listening is active and analytical: it asks "what does this tell us?" rather than simply recording that it happened.
Social listening is not the same as social media reporting. Reporting is quantitative — it measures reach, engagement rate, follower growth, and return on investment using platform analytics. Listening is qualitative: it reads the content of what people say, not just how many people said it. A business can have excellent reporting numbers and still be completely unaware that a forum thread is circulating a complaint about their customer service. Listening catches what the metrics miss. Do not conflate the two; they serve different purposes and require different skills.
For clients in Uganda and East Africa, listening carries a particular strategic weight. The majority of customer feedback in this market travels through WhatsApp — a private, end-to-end encrypted channel that cannot be monitored externally. What reaches public platforms — Facebook comments, Google reviews, X/Twitter posts — is therefore a fraction of the total conversation. That fraction is nonetheless a meaningful proxy: if five customers complain publicly on Facebook, it is reasonable to infer that many more have shared the same frustration privately. Listening catches the visible signal and uses it to understand the invisible majority. The EA consultant who monitors consistently will always know more about their client's reputation than one who relies on the client to report problems themselves.
Build the client's keyword list across five categories before setting up any tool. A complete taxonomy ensures no critical mentions are missed and avoids wasting time monitoring irrelevant noise.
| Keyword type | Example | Client's version |
|---|---|---|
| Exact business name | "Kampala Fresh Bakery" | |
| Common misspelling 1 | "Kampala Fresh Bakary" | |
| Common misspelling 2 | "Kla Fresh Bakery" | |
| Product / service name | "sourdough bread Kampala" | |
| Founder name (if public) | "Sarah Nakato bakery" | |
| Brand hashtag | #KampalaFreshBakery |
| Keyword type | Example | Client's version |
|---|---|---|
| Competitor 1 brand name | "City Breads Uganda" | |
| Competitor 2 brand name | "Kampala Cakes" | |
| Competitor 3 brand name | "Bake House UG" | |
| Competitor product name | "City Breads croissant" | |
| Comparison search | "vs City Breads" |
| Keyword type | Example | Client's version |
|---|---|---|
| Customer-language category term | "fresh bread Kampala" | |
| Common customer question | "where to buy sourdough Kampala" | |
| Industry hashtag | #KampalaBakery | |
| Discovery phrase | "best bakery in Kampala" |
| Sentiment | Example search string | Client's version |
|---|---|---|
| Negative | "[Brand] complaint" | |
| Negative | "[Brand] bad / scam / disappointed / problem" | |
| Positive | "[Brand] recommend / excellent / love / happy" | |
| Positive | "[Brand] best" |
| Keyword type | Example | Client's version |
|---|---|---|
| Brand + city | "[Brand] Kampala" | |
| Brand + country | "[Brand] Uganda" | |
| Brand + region | "[Brand] East Africa" | |
| Industry + city (discovery) | "bakery Kampala" | |
| Local language brand variant | "omugati" (bread in Luganda) |
Work through this table with the client at onboarding. A complete taxonomy takes 20–30 minutes to build and saves hours of wasted monitoring time.
Set up all four tools below. Each covers a different source type; together they provide sufficient coverage for an EA small-to-medium business without paid subscription costs.
Google Alerts monitors news sites, blogs, and the open web — not social media. It is the best free tool for catching press coverage, blog mentions, and forum posts.
Setup steps:
Limitation: Google Alerts does not index Facebook, Instagram, TikTok, or WhatsApp. It is a supplement to platform monitoring, not a replacement.
Conduct manual searches on each platform weekly. This is the most reliable way to find social mentions; it requires no tool, no account upgrade, and no budget.
Facebook:
Instagram:
TikTok:
X / Twitter:
LinkedIn:
Use Brand24 (brand24.com) or Mention (mention.com) for automated mention tracking. Both offer free tiers adequate for small clients starting out.
Brand24 free tier: monitors up to 3 keywords; shows a limited number of recent mentions. Set up: brand name, primary competitor name, primary industry keyword.
Mention free tier: 250 mentions per month. Set up the same 3 keywords.
Setup steps (Brand24 example):
Budget note: Paid tiers for Brand24 start at approximately $49 per month (approximately UGX 180,000). Flag this to the client. For most small EA clients, the free tier combined with native platform search is sufficient until monthly social media revenue warrants the upgrade.
GBP reviews are one of the highest-intent feedback sources available — customers who leave a review have had a direct experience with the business.
Setup steps:
Build listening into the weekly workflow. Unscheduled monitoring does not happen consistently.
meta-competitor-analysisRaw mentions have no value until they are interpreted. Apply these five questions to listening data every week.
The 5 Weekly Listening Questions:
What are customers praising about us this week?
Surface these in the log as "amplify" opportunities. Praised experiences become content: a customer complimenting the bakery's sourdough is a prompt for a Reel about the sourdough-making process. Share the finding with the content calendar (see 11-content-calendar).
What complaints or frustrations appeared? Log each complaint with its platform, sentiment rating, and whether it was resolved. Recurring complaints (the same theme appearing three or more times across separate posts) indicate a service or product issue — escalate to the client's operations team, not just the social media manager.
What are people saying about our competitors?
Competitor complaints are positioning opportunities. If three customers in one week complain that a competitor's delivery is slow, and the client offers faster delivery, that is a campaign angle. Feed findings into 09-campaign-strategy.
What questions are people asking about our category?
Unanswered category questions are content gaps. A cluster of "where can I find sourdough bread in Kampala?" searches means no one is answering that question well. The client who answers it first wins the discovery. Feed findings into 11-content-calendar as FAQ content.
Are there any emerging topics in our industry we should comment on?
New regulations, food safety news, industry trends, local events — topics gaining traction in listening are thought leadership opportunities. Feed findings into 10-content-pillars if a theme is recurring.
Weekly debrief format: Complete the five questions above in writing. Takes 15 minutes. Share the written debrief with the client once per month, collated as a monthly listening summary. This makes the intelligence visible and justifies the monitoring time investment.
Add these fields to every material listening entry: context, audience, public/private, interaction type, confidence, action owner, decision changed, and guardrail or escalation. If the source is inaccessible, record the check as not assessed rather than filling the gap with a conclusion.
Maintain this log as a Google Sheet. Create one tab per month. Share view access with the client.
| Date | Platform | Mention type | Sentiment | Summary (1 sentence) | Action taken | Priority |
|---|---|---|---|---|---|---|
| DD/MM/YYYY | Facebook / Instagram / GBP / TikTok / X / Other | Brand / Competitor / Industry | Positive / Neutral / Negative | High / Medium / Low |
Column guidance:
Example rows (Kampala business — bakery):
| Date | Platform | Mention type | Sentiment | Summary | Action taken | Priority |
|---|---|---|---|---|---|---|
| 15/03/2026 | Brand | Negative | Customer posted that their birthday cake order arrived 2 hours late with no apology from staff | Replied publicly; escalated to client; issued DM apology with voucher | High | |
| 17/03/2026 | Google Business Profile | Competitor | Negative | 3 reviews this week on City Breads profile complaining about dry cake texture | Added "freshness guarantee" angle to next month's content brief | Medium |
| 18/03/2026 | TikTok | Industry | Positive | Creator with 12K followers posted taste-test video of Kampala bakeries; client not featured | Identified creator; flagged to client for potential collaboration | Medium |
Listening is the intelligence layer that makes every other skill in this suite more effective. Use the findings actively — do not let them sit in a log.
→ Content calendar (11-content-calendar):
Themes and questions surfaced in listening become content ideas for the following month. Add listening-derived ideas to the content brief at the monthly planning session.
→ Content pillars (10-content-pillars):
If the same theme appears in listening for three or more consecutive months, review whether it warrants its own content pillar or an adjustment to an existing one. Listening validates whether the current pillar structure reflects what the audience actually cares about.
→ Campaign strategy (09-campaign-strategy):
Competitor weaknesses identified in listening become campaign angles. A documented pattern of competitor complaints is evidence for a positioning campaign — "We do what [competitor] gets wrong."
→ Community management (playbook-community-management):
Unresolved complaints found in listening that were not responded to on the platform become community management priorities. Check the log at the start of each community management session.
→ Crisis communications (playbook-crisis-communications):
Listening is the early warning system for reputational risk. A sudden increase in negative brand mentions — particularly if the same topic appears across multiple platforms in the same 48-hour window — is a Level 1 crisis trigger. Escalate immediately using the crisis classification in playbook-crisis-communications.
→ Product and service feedback (client operations): Listening intelligence about product frustrations, delivery problems, and service failures has operational value beyond social media. Share a monthly summary of product-related complaints with the client's operations lead. Social listening is a low-cost substitute for formal customer research in markets where research budgets are limited.
Multilingual monitoring: Ugandan customers regularly mix English and Luganda in social media posts — this is called Luganda-English code-switching and is normal in urban Kampala. Standard keyword monitoring will miss Luganda-language mentions entirely unless you build them in. Work with the client at onboarding to identify the top 5 product and service names as customers actually say them in conversation. Examples: "omugati" (bread), "emmere" (food), "ssente" (money/payment), "omusawo" (doctor). Add these as separate keywords in Brand24 and in platform searches. Apply the same principle for Swahili in the Kenyan or Tanzanian market and Sheng for urban Nairobi audiences.
WhatsApp is unmonitorable — acknowledge this explicitly: WhatsApp is the primary channel for customer feedback in Uganda, and it is end-to-end encrypted. No external tool can monitor it. Do not imply otherwise. What can be done: brief the client's customer-facing staff to flag recurring WhatsApp complaints themes monthly. Create a simple staff feedback form (a shared Google Sheet or WhatsApp Group poll) asking: "What are the top 3 complaints or questions you received via WhatsApp this month?" Incorporate the responses into the monthly listening debrief.
Facebook Groups: A significant proportion of EA community and industry conversation happens in private or semi-private Facebook Groups rather than on public pages. Identify 3–5 relevant groups at onboarding — local industry groups, neighbourhood community groups, city-specific consumer groups (e.g., "Kampala Foodies", "Kampala Business Network"). Join them using an agency or client account and monitor manually each week. Note that Facebook Group content does not appear in external tool searches.
Informal and tabloid media monitoring: In Uganda, informal online news outlets such as Sqoop (sqoop.co.ug), Nile Post (nilepost.co.ug), and Chimp Reports (chimpreports.com) can amplify negative stories quickly and are read widely. Add their RSS feeds to Google Alerts or Feedly (free) so that coverage appears in the daily digest. Do not rely solely on mainstream media monitoring.
Output meets the standard if it: