Use this skill whenever a user needs help writing for non-academic audiences or communicating research to the public. Triggers include: any mention of "op-ed," "blog post," "public writing," "writing for the public," "Sapiens," "The Conversation," "public anthropology," "popular writing," "community report," "return of results," "podcast preparation," "media interview," "talking points," "press release," "public scholarship," "engaged anthropology," "translating research," "accessible writing," "plain language," "policy brief," "community summary," or "media training." Covers public-facing writing (blog posts, op-eds, popular press), podcast and media preparation, community summaries and return-of-results documents, and register translation from academic to accessible language. Do NOT use for academic conference presentations (use conference-materials skill), peer-reviewed article writing, or grant proposals (use grant-proposal skill).
From ai-anthropologynpx claudepluginhub mattartzanthro/ai-anthropology-toolkit --plugin ai-anthropologyThis skill uses the workspace's default tool permissions.
references/community-engagement-guide.mdreferences/media-preparation-guide.mdreferences/public-writing-guide.mdGuides Next.js Cache Components and Partial Prerendering (PPR) with cacheComponents enabled. Implements 'use cache', cacheLife(), cacheTag(), revalidateTag(), static/dynamic optimization, and cache debugging.
Migrates code, prompts, and API calls from Claude Sonnet 4.0/4.5 or Opus 4.1 to Opus 4.5, updating model strings on Anthropic, AWS, GCP, Azure platforms.
Reviews prose for communication issues impeding comprehension, outputs minimal fixes in a three-column table per Microsoft Writing Style Guide. Useful for 'review prose' or 'improve prose' requests.
Help anthropologists communicate research to non-academic audiences across formats and platforms. This skill treats public engagement as register translation — converting academic knowledge into accessible communication without oversimplification or loss of nuance. The goal is not to strip away complexity but to repackage it for audiences who have not spent years in your subfield's conversations.
Public engagement is not "dumbing down" — it is a different rhetorical genre with its own craft requirements. Op-eds need a clear take, not a balanced literature review. Community reports need actionable findings, not hedged conclusions. Podcast prep needs soundbites that don't distort. Each format demands different skills, and the failure modes are different from academic writing. The most common mistake is treating public writing as academic writing with the jargon removed — it is not. It has its own architectures, its own narrative logics, and its own standards of evidence presentation.
| Task | Reference |
|---|---|
| Op-eds, blog posts, popular articles, register translation | Read references/public-writing-guide.md |
| Community reports, return-of-results, reciprocity documents | Read references/community-engagement-guide.md |
| Podcast, radio, media interviews, talking points | Read references/media-preparation-guide.md |
Determine the entry point. Public engagement spans several distinct genres, and the user may need one or several:
Ask the user if not immediately clear:
Collect the information needed to produce effective public-facing content. Not all of this is needed upfront — gather what you can and note gaps for the user to fill.
Essential context (cannot proceed without these):
Important context (strengthens the output significantly):
Helpful context (improves tailoring):
references/public-writing-guide.md for platform
formats, narrative structures, register translation principles, pitching
strategies, and ethical considerations.references/community-engagement-guide.md for
reciprocity frameworks, report structures, consent boundaries, plain
language principles, and community review processes.references/media-preparation-guide.md for talking point development,
interview techniques, soundbite crafting, and post-interview strategies.Follow the audience-first principle across all formats:
For public writing:
For community reports:
For media preparation:
For register translation:
Produce one or more of these deliverables depending on user needs:
Before presenting the output, verify:
| Failure mode | Prevention |
|---|---|
| Academic tone in public writing — passive voice, hedging, literature review structure transplanted into a blog post | Rewrite in active voice, first person where appropriate; lead with the finding or story, not the gap in the literature |
| Jargon without definition — "neoliberal governmentality," "ontological turn," "biopolitics" dropped without explanation | Define or replace every technical term; if a term is essential, define it in plain language on first use |
| No clear stake — the piece describes research but never explains why anyone outside the discipline should care | Require a "so what" statement in the first or second paragraph; every public piece must answer "why does this matter to you?" |
| Community report that serves the researcher — reports what the researcher finds interesting rather than what the community needs to know | Ask explicitly what community partners want to learn; structure around community questions, not research questions |
| Oversimplification that distorts — a complex finding flattened into a misleading headline or soundbite | Flag when simplification crosses into distortion; offer the user a "nuanced version" and a "simplified version with caveats" |
| Consent boundary violation — sharing stories, quotes, or identifiable details that participants did not consent to have made public | Check consent scope before drafting; default to more restrictive interpretation; composite or anonymized examples when needed |
| Reading from a script in media — talking points treated as a script rather than a preparation framework, resulting in stilted delivery | Frame talking points as key messages to internalize, not text to recite; include practice exercises for conversational delivery |
| Missing the news hook — pitching a timely piece too late, or writing an op-ed that has no connection to current events or debates | Identify the news hook or public conversation the piece enters; if no hook exists, consider whether a blog or feature is a better format |
Example 1: Op-ed for a newspaper opinion page
Input: "I want to write an op-ed about my research on maternal health in rural Guatemala. There's a new WHO report on maternal mortality that just came out, and I think my fieldwork findings complicate the report's recommendations. I've been doing ethnographic research for three years in two highland communities."
Output approach: Load the public writing guide. Set parameters: audience = general public; platform = op-ed (newspaper); purpose = public scholarship with advocacy dimensions; register = general public. Identify the news hook — the WHO report provides a timely entry point. Structure the op-ed (~700 words): open with a vivid fieldwork scene that illustrates the gap between policy recommendations and lived reality (not "my research examines" — show, don't tell). Second paragraph: the WHO report and its recommendations. Third paragraph: what the report gets wrong or misses, grounded in ethnographic evidence. Fourth paragraph: why this matters — the real consequences of policy that does not account for local realities (specific examples from fieldwork, with identities protected). Fifth paragraph: what should be done differently — a concrete recommendation or reframing. Close with a return to the human story. Include a suggested headline, an author bio line, and a note on which outlets to pitch to (newspapers with health or global affairs coverage, or outlets like The Guardian's global development section). Flag: verify that any patient stories or community details shared in the op-ed fall within consent boundaries for public dissemination.
Example 2: Community report and return of results
Input: "I've finished a two-year CBPR project with a Somali immigrant community in Minneapolis studying barriers to healthcare access. The community advisory board wants a report they can share with community members and use in advocacy with the city health department. I also need a shorter version for the health department directly."
Output approach: Load the community engagement guide. Set parameters: audience = community members (primary), policymakers (secondary); platform = community report (full) + policy brief (short version); purpose = reciprocity and advocacy; register = general public for the community report, policymaker for the policy brief. For the community report: structure around the questions the community advisory board identified as priorities (not the researcher's academic research questions). Executive summary in plain language (one page). Key findings presented with community voice — include participant quotes (with consent), community advisory board interpretations, and space for disagreement or additional context. Implications section focused on what the community can do with the findings. Next steps section co-developed with the advisory board. Accessible formatting: large font, visual aids, translated sections if needed (confirm languages). For the policy brief: 2-3 pages, executive summary with three bullet-point recommendations, evidence summary linking each recommendation to specific findings, implementation considerations. Formal but accessible register appropriate for city health department staff. Both documents should be reviewed by the community advisory board before distribution — include a note about the review process and timeline.
Example 3: Podcast preparation and talking points
Input: "I've been invited on a popular science podcast to talk about my research on how conspiracy theories spread in online communities. The host is a science journalist and the audience is educated non-specialists. The episode is next week and I'm nervous — I've never done a podcast before. My research is based on 18 months of digital ethnography."
Output approach: Load the media preparation guide. Set parameters: audience = educated non-specialists; platform = podcast; purpose = knowledge sharing / public scholarship; register = educated non-specialist (can handle complexity, but not disciplinary vocabulary). Develop three key messages: (1) what the researcher actually found (the surprising or counterintuitive finding that will hook listeners), (2) why it matters beyond the academic context (implications for how we think about misinformation, public health, democratic participation), (3) what the researcher's approach reveals that other approaches miss (the value of ethnographic attention to meaning-making rather than just information flow). For each key message, craft one strong soundbite — a memorable, accurate, quotable statement that could be pulled for a clip or social media promotion. Prepare anticipated questions: "Aren't conspiracy theorists just irrational?" (bridge to complexity of belief systems), "What should platforms do about this?" (bridge to what the research shows vs. policy prescriptions), "How did you study this?" (opportunity to explain digital ethnography accessibly). Include bridge phrases for redirecting away from oversimplification ("That's a common assumption, but what I found is...") and strategies for the "well actually" trap (avoid lecturing; tell stories instead). Provide a one-paragraph elevator pitch for the research, and a pre-interview checklist: research the host's style, listen to two previous episodes, identify the audience's likely prior knowledge, prepare an opening anecdote. Note first-time podcast advice: speak in shorter sentences than feels natural, pause before answering, it is fine to say "that's a great question, let me think about that."