strategic-narrative-generator
Skill13 starsChanged 5 months agoMITView on GitHub
Generates the strategic story connecting your roadmap to company
In this repository23 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- ambiguity-resolverSkill
- challengeSkill
- deep-thinkSkill
- design-critique-templateSkill
- experiment-designSkill
- ingestSkill
- interview-analysisSkill
- multi-source-signal-synthesiserSkill
- next-advisorSkill
- pipeline-stepsSkill
- post-launch-reviewSkill
- problem-structuringSkill
- product-requirements-docSkill
- setup-initiativeSkill
- solution-scoringSkill
- system-design-docSkill
- technical-spec-documentSkill
- tracking-and-funnelsSkill
- ui-pattern-librarySkill
- user-persona-builderSkill
- user-story-generatorSkill
- user-testingSkill
The file
We split this skill into the rules it gives an agent
9 rules. About 0 are common, 7 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 31
Read the initiative list and identify 2-3 natural strategic themes
- Distinctive: line 35
Write executive summary in under 100 words (the version someone can repeat)
- Distinctive: line 71
Write for a CFO, not an engineer
- Distinctive: line 73
Every sentence should answer "so what?"
- Distinctive: line 74
Avoid jargon — if you can't say it plainly, the strategy isn't clear enough yet
- Distinctive: line 183
Every claim in a Gate presentation must include a source:
- Very specific: line 185
Evidence typing: If the claim is based on SYNTHETIC data, mark it: [Synthetic — needs validation]
- Distinctive: line 186
No unsourced claims: If you can't trace a claim to a file, either find the source or remove the claim
- Very specific: line 198
Run the anti-generic checklist from .claude/rules/output-formats.md (§ Anti-generic self-check): no unsourced claims, no placeholder …
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Strategic Narrative Generator Skill
Purpose
Turn a prioritised initiative list into a strategic narrative — the story that explains not just what you're building but why, why now, and why this sequence. The kind of narrative a board member can repeat back correctly after one hearing.
Also structures Gate presentation narratives where every claim must be backed by a specific source from research and output artifacts.
Required Inputs
- Prioritised initiative list (with rough timelines)
- Current OKRs or strategic priorities (1-3)
- Competitive or market context (optional but improves output significantly)
Process
- Read the initiative list and identify 2-3 natural strategic themes
- For each theme: articulate the problem it addresses, the customer it serves, and the metric it moves
- Build the progression narrative: how does Q1 set up Q2? How does H1 set up H2?
- Write executive summary in under 100 words (the version someone can repeat)
- Anticipate the 3 hardest questions a sceptical board member would ask — and draft answers
- Identify what's NOT on the roadmap and why (this builds credibility)
Output Format
Product Strategy Narrative: [Period]
The One-Paragraph Context: [Market moment + key challenge + our response — for the CFO, not the engineer]
Strategic Theme 1: [Name]
- The problem: [customer pain in plain language]
- Our response: [initiatives in this theme]
- The metric it moves: [specific and measurable]
- Why now: [timing rationale]
Strategic Theme 2: [Name] [Same structure]
The Progression Story: [How each quarter sets up the next — this is the narrative arc]
Executive Summary (under 100 words — shareable): [Version someone can quote at a board meeting]
Questions to Prepare For:
- [Hard question] -> [Prepared answer]
- [Hard question] -> [Prepared answer]
- [Hard question] -> [Prepared answer]
What's Not on the Roadmap (and Why): [2-3 items — shows strategic discipline, not just prioritisation]
Tone Rules
- Write for a CFO, not an engineer
- Lead with outcomes, not features
- Every sentence should answer "so what?"
- Avoid jargon — if you can't say it plainly, the strategy isn't clear enough yet
Report Presentation Narratives (Steps 10 and 15)
Problem Research Report Structure (Step 10: /create-presentation)
Purpose: Convince stakeholders the problem is real, validated, and worth solving.
## Slide 1: Title
[Initiative name + PM name + date]
## Slide 2: Context
- Where this task comes from (OKR, user feedback, data signal)
- Why now (what changed)
- **Source**: CONTEXT.md — "Why now" field
## Slide 3: Problem
- Problem thesis (one sentence)
- Who's affected: segment + size
- Signal strength: how many sources confirm this
- **Sources**: hypotheses.json (registry — statuses/confidence), hypotheses.md/validated-hypotheses.md (narrative)
## Slide 4: AS IS Scenario
- Current user behavior (from CJM analysis)
- Pain points with evidence
- Quotes from research (if REAL) or synthetic insights (marked)
- **Sources**: CJM/ materials, synthetic-interviews.md, interview-notes.md
## Slide 5: Hypothesis
- Formula: "If [X], then [Y], because [Z], metric [M] +[N%]"
- Confidence level with evidence type breakdown
- **Sources**: solution-hypotheses.md
## Slide 6: Solution
- Job-to-be-done
- Key screens / user flow (from solution-sketch.md)
- Use cases: 2-3 concrete scenarios
- **Sources**: solution-sketch.md, competitive-analysis.md
## Slide 7: Estimate
- Timeline and effort
- Key risks and dependencies
- What we need to proceed (Gate decision)
- **Sources**: CONTEXT.md constraints, dev-estimate.md (if available)Speaker notes template for each slide:
### Speaker Notes — Slide [N]: [Title]
**Key message** (say this first): [One sentence the audience should remember]
**Supporting points**: [2-3 bullets to elaborate]
**Anticipated question**: [What someone might ask here]
**Answer**: [Prepared response with source]
**Transition**: [How to move to next slide]Solution Research Report Structure (Step 15: /create-gate2-presentation)
Purpose: Convince stakeholders the solution is ready for development and testing.
## Slide 1: Title
[Initiative name + PM name + date]
## Slide 2: Hypothesis Recap
- Formula from Problem Research Report
- Metrics: primary + guardrail + proxy
- Target audience and size
- **Sources**: solution-hypotheses.md, PRD.md §3
## Slide 3: Solution Context (AS IS)
- Current user journey + pain points (recap)
- What changed since Problem Research Report (new data, design feedback)
- **Sources**: hypotheses.json (registry), concept-test-results.md
## Slide 4: Solution Design
- Job-to-be-done
- Key screens / mockups (from designer or wireframes)
- User flow
- **Sources**: solution-sketch.md, design-brief.md
## Slide 5: Demo
- Live demo or clickthrough of key screens
- [PM presents this live — slides as backup]
## Slides 6-7: UX Test Results (if conducted)
- Methodology: who, how many, how
- Key findings: what worked, what didn't
- Changes made based on feedback
- **Sources**: concept-test-results.md, ux-research-brief.md
## Slide 8: Experiment Design
- AB test parameters: baseline, MDE, sample, duration
- Guardrail metrics
- Decision criteria: ship / iterate / kill
- **Sources**: ab-test-design.md
## Slide 9: Estimate & Timeline
- Dev effort (from dev lead)
- Dependencies and risks
- Proposed timeline: development -> AB test -> decision
- **Sources**: dev-estimate.md, PRD.md §9-§10Source Reference Rules
Every claim in a Gate presentation must include a source:
- Format:
[Source: filename.md]or[Source: filename.md, section] - Evidence typing: If the claim is based on SYNTHETIC data, mark it:
[Synthetic — needs validation] - No unsourced claims: If you can't trace a claim to a file, either find the source or remove the claim
- Qualitative data (interviews, synthetic) is illustration only — don't present as proof without quantitative backing
References (read when assembling a gate deck)
references/mckinsey.md— storylining, horizontal/vertical logic, slide craft. Read when structuring Problem/Solution Research Report decks.references/exec-communication.md— executive audience rules, answer-first delivery, objection handling. Read before high-stakes presentations.
Mandatory self-check before delivering
Run the anti-generic checklist from .claude/rules/output-formats.md
(§ Anti-generic self-check): no unsourced claims, no placeholder advice,
proof slides REAL-only, frequency honesty ("N of M sources"), numbers
cross-checked against the hypothesis registry. Fix violations before
showing the draft — never deliver and apologize later.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

