The CO-STAR Framework
The CO-STAR Framework is designed for context-heavy communication where rich background information is essential for quality outputs. It's particularly effective for reports, summaries, and strategic documents.
Framework Components
CO-STAR is an acronym for:
| Component | What It Means | Why It Matters |
|---|---|---|
| C | Context | Background information and situation |
| O | Objective | The goal you want to achieve |
| S | Style | Writing style and approach |
| T | Tone | Voice and emotional quality |
| A | Audience | Who will read/use this output |
| R | Response | Format and structure of output |
Component Breakdown
C – Context
Provide rich background information.
This includes: - Current situation - Why this task is needed - Relevant constraints - Historical background - Related work
Example:
Context: AmaDema is preparing for a Series A funding round focused
on sustainable materials. We need to demonstrate our non-oxide
ceramic nanofibers' environmental advantages over traditional
polymer competitors. Our target investors prioritise ESG
(Environmental, Social, Governance) metrics. We have strong
technical data but lack compelling narrative for non-technical
stakeholders.
O – Objective
Define the specific goal.
Be explicit about what success looks like:
❌ "Write about our materials"
✅ "Create a compelling narrative that positions our non-oxide ceramics as the sustainable alternative to PLA, emphasising lifecycle carbon footprint reduction of 40%"
S – Style
Define the writing approach.
Common styles:
- Academic: Structured, cited, peer-review quality
- Journalistic: Engaging, story-driven, accessible
- Technical: Precise, data-rich, domain-specific
- Executive: High-level, business-focused, concise
- Conversational: Friendly, approachable, informal
Example:
Style: Journalistic with technical credibility—engaging narrative
for non-scientists, but backed by quantitative data. Think
"Nature News" rather than primary research article.
T – Tone
Set the emotional quality and voice.
Scientific tones:
- Confident: Asserting leadership position
- Cautious: Emphasising uncertainty, safety
- Inspirational: Vision-driven, future-focused
- Objective: Neutral, evidence-based
- Urgent: Emphasising timeliness, action needed
Example:
Tone: Confident but not arrogant. Emphasise innovation and
environmental responsibility. Avoid hype—let data speak.
Inspire confidence in our technical capability.
A – Audience
Define who will consume this.
Key questions:
- What is their technical background?
- What are their priorities/concerns?
- What do they already know?
- What terminology level is appropriate?
Example:
Audience: Venture capital investors with general business
background (not materials science experts). Familiar with
sustainability concepts but need help understanding technical
differentiation. Sceptical of greenwashing—want hard data.
R – Response Format
Specify structure and length.
Examples:
Response: 2-page executive summary with:
- Opening hook (1 paragraph)
- Problem statement (2 paragraphs)
- Our solution (3 paragraphs)
- Market opportunity (2 paragraphs)
- Competitive advantage (2 paragraphs)
- Call to action (1 paragraph)
Include 2-3 data visualisation callouts (we'll create separately).
Complete CO-STAR Example
Scenario: Investor Pitch Document
Prompt:
[C] Context: AmaDema is a 3-year-old nanotechnology startup
specialising in non-oxide ceramic nanofibers produced via
electrospinning. We're preparing for Series A funding ($5M target)
focused on scaling our sustainable materials platform. Our core
innovation is a novel synthesis route that reduces carbon footprint
by 40% vs. traditional PLA while maintaining comparable mechanical
properties. We have IP protection (2 granted patents, 3 pending),
pilot-scale production facility (500kg/month capacity), and 3
industry partners (aerospace, automotive, biomedical). Competitors
include PolyMat GmbH (Germany, polymer focus) and Nanotech Solutions
(US, oxide ceramics only). Our target investors prioritise ESG
metrics and deep-tech with clear commercialisation path.
[O] Objective: Create an executive summary for our pitch deck that
positions AmaDema as the leading sustainable alternative to polymer
nanofibers, emphasising our environmental advantages, technical
moat, and clear path to profitability. Goal is to secure meetings
with 15 VC firms over next quarter.
[S] Style: Professional but engaging. Think TechCrunch feature
article meets technical white paper executive summary. Use
storytelling to make technical innovation accessible, but back
every claim with quantitative data. Structure should flow naturally
while hitting all key investment criteria (market, technology,
team, traction, ask).
[T] Tone: Confident and visionary, but grounded in evidence.
Emphasise "inevitable future" of sustainable materials rather than
speculative opportunity. Inspire FOMO (fear of missing out) on
a transformative materials platform. Avoid hype language—let metrics
do the talking. Professional but not stuffy.
[A] Audience: Venture capital partners and principals at deep-tech
focused firms (e.g., DCVC, Lux Capital, Breakthrough Energy
Ventures). They understand business fundamentals and sustainability
trends, but not materials science details. Familiar with terms like
"carbon footprint" and "lifecycle analysis" but need help
understanding why non-oxide ceramics are technically superior.
Sceptical of greenwashing—require hard data. Decision criteria:
market size, defensibility, team quality, capital efficiency, exit
potential.
[R] Response: 2-page executive summary (approximately 800 words) with:
Section 1: Opening Hook (1 paragraph)
- Lead with compelling problem/opportunity
- Immediately establish market urgency
Section 2: The Problem (2 short paragraphs)
- Why current polymer nanofibers inadequate
- Environmental and performance limitations
Section 3: Our Solution (3 paragraphs)
- Non-oxide ceramic nanofibers overview
- Key technical innovation (synthesis route)
- Quantitative advantages (carbon footprint, properties)
Section 4: Market Opportunity (2 paragraphs)
- TAM/SAM/SOM framework
- Target applications (aerospace, automotive, biomedical)
Section 5: Competitive Advantage & IP Moat (2 paragraphs)
- What makes us defensible
- Patent portfolio and technical barriers
Section 6: Traction & Milestones (1-2 paragraphs)
- Current partnerships
- Production capacity
- Recent achievements
Section 7: The Ask (1 paragraph)
- Funding amount and use of funds
- Next milestones
- Clear call to action
Include [DATA VISUALISATION CALLOUT] markers where we should insert
charts (we'll create separately).
Use bold for key metrics. Keep paragraphs under 100 words.
AUTOMAT vs. CO-STAR: When to Use Each
Use AUTOMAT When:
✅ Task is functional and structured (data extraction, formatting)
✅ Output format is clearly defined (tables, lists, code)
✅ Precision and reproducibility are critical
✅ Minimal context needed
Example tasks: Literature data extraction, protocol formatting, data analysis scripts
Use CO-STAR When:
✅ Task requires rich contextual understanding
✅ Output is narrative or strategic (reports, pitches, summaries)
✅ Audience considerations are complex
✅ Style and tone significantly impact effectiveness
Example tasks: Executive summaries, investor pitches, strategic reports, literature reviews with interpretation
Materials Science Example: Literature Review
Using CO-STAR for Research Synthesis
[C] Context: I'm preparing the introduction section for a manuscript
on PLA/graphene nanocomposites being submitted to Polymer journal.
The field is crowded with incremental studies, but most fail to
address the scalability challenges of graphene dispersion. Our work
introduces a novel in-situ reduction approach during electrospinning
that achieves uniform dispersion at industrial-relevant concentrations
(>5 wt%). Reviewers will be experts in polymer nanocomposites. The
introduction needs to establish the research gap that justifies our
study while acknowledging prior work comprehensively.
[O] Objective: Create a literature review synthesis for the
introduction (approximately 1000 words) that: (1) establishes the
promise of PLA/graphene composites, (2) identifies the critical
challenge of graphene dispersion at high loadings, (3) reviews
existing approaches and their limitations, (4) sets up our in-situ
reduction approach as the logical next step.
[S] Style: Academic, suitable for peer-reviewed journal. Follow
Polymer journal's convention of critical synthesis rather than
exhaustive listing. Use topic sentences to guide reader through
logical argument. Balance breadth (covering field) with depth
(analysing key studies). Integrate citations naturally into prose
rather than citation-heavy listing.
[T] Tone: Authoritative but respectful of prior work. Identify gaps
without dismissing others' contributions. Build logical case for
our approach without over-selling. Objective analysis of field state.
[A] Audience: Polymer scientists and materials engineers with
expertise in nanocomposites. Assume familiarity with electrospinning
and graphene chemistry. No need to define basic terms (PLA,
graphene oxide, in-situ reduction), but explain domain-specific
methods. Reviewers will fact-check citations—accuracy critical.
[R] Response: 4-paragraph structure (approximately 250 words each):
Paragraph 1: Promise of PLA/graphene
- Mechanical property enhancement potential
- Sustainability advantages
- Application opportunities
Paragraph 2: The dispersion challenge
- Aggregation at high loadings
- Impact on properties
- Why this limits commercialisation
Paragraph 3: Existing approaches and limitations
- Ex-situ methods (sonication, surfactants)
- In-situ polymerisation approaches
- Critical analysis of each
Paragraph 4: Gap and our approach
- What's missing: scalable in-situ method
- Preview our solution (without methods details)
- Research questions
Cite 30-40 key papers, prioritising recent (<5 years) and
high-impact journals (Polymer, Composites A, Carbon). Mark citation
format as [Author Year] for now.
Combining AUTOMAT and CO-STAR
For complex tasks, you can use both frameworks sequentially:
Step 1: Use CO-STAR for narrative/synthesis
Generate the prose, interpretation, or strategic content
Step 2: Use AUTOMAT for data extraction/formatting
Extract structured data to support the narrative
Example workflow:
- CO-STAR prompt: "Write executive summary of our sustainability advantage"
- Review output: Identify where specific data would strengthen argument
- AUTOMAT prompt: "Extract lifecycle carbon footprint data from these 5 papers in a comparison table"
- Integrate: Insert table into executive summary
Exercise: CO-STAR for Strategic Communication
Challenge
Scenario: AmaDema's R&D Director asked you to write a brief (1-page) internal memo explaining why the team should adopt AI tools for literature review, addressing common concerns about accuracy and IP security.
Your task: Construct a complete CO-STAR prompt for this task.
Consider: - What context does the R&D Director need? - What's your objective beyond "explain AI tools"? - What style/tone will be persuasive for scientists? - How should you structure the response?
Test in sandbox and evaluate whether output would be convincing to your colleagues.
Common Mistakes to Avoid
❌ Mistake 1: Context Overload
Bad:
Good:
Rule: Include context relevant to this output, not your entire company history.
❌ Mistake 2: Vague Objective
Bad:
Good:
Objective: Position our non-oxide ceramics as the sustainable
alternative that justifies 20% price premium by emphasising
lifecycle cost savings and regulatory advantages
❌ Mistake 3: Style/Tone Confusion
Bad:
Good:
Style: Academic synthesis with critical analysis (Polymer journal
conventions)
Tone: Authoritative but respectful of prior work; objective analysis
rather than promotional
Next: Excercises →