You ARE the bottleneck.You know it.
The board memo is going to be a panic-write again. The investor update is six weeks late. You can’t remember what the last 12 customers told you. Runway feels close but you haven’t modeled it since the Series A. Monday morning is a fresh fog of nothing-planned.
The Solo Founder Skills Pack is six Claude skills that load themselves when you need them. Built on the May 2026 Agent Skills format. Deployable in 20 minutes. The second brain you’ve been promising yourself for two years.
You bought every founder productivity tool. None of them solved the actual problem.
You have a ChatGPT tab open. A Claude tab open. Three custom GPTs you built once and never opened again. A Notion graveyard of half-built systems. A prompt pack PDF buried in Downloads. None of them actually load themselves when you need them. None of them remember what you said last month. None of them feed each other. You are still copy-pasting, still re-explaining, still being the glue that holds everything together — which means you are still the bottleneck.
The Solo Founder Skills Pack is built on a different premise: you should not have to know which tool to use. Claude should recognize the context, load the right skill, and produce the right output — every time, in any environment, with your business’s vocabulary baked in. That’s what the May 2026 Agent Skills format makes possible. That’s what this pack delivers, pre-built, for the six tasks every solo founder does and nobody does well.
Average lateness of the next investor update on every solo founder's calendar. The skill makes it a 20-minute task.
The number of functions a solo founder does badly because they're doing seven of them. The pack handles the structural part of six.
Average deployment time to install the pack into Claude Pro or Claude Code. Five environments supported.
Clear about the lane. No inflated promises.
- Six Claude skills built on the May 2026 Agent Skills open standard.
- Auto-loading — Claude detects context and invokes the right skill.
- Composable — skills feed each other (interview synth → update writer).
- Portable — works in Claude Pro, Claude Code, Codex CLI, Cursor, Gemini CLI, Copilot.
- Quality-tuned for real fundraising rounds and real board meetings.
- 12 worked examples, 12 months of template updates.
- A prompt pack. Prompts are passive copy-paste. Skills are active triggers.
- A custom GPT. Custom GPTs don't compose, don't auto-load, don't port.
- A Notion template. Templates need you to fill them. Skills produce them.
- A SaaS subscription. One-time $79.
- A CEO replacement. The pack structures your thinking; you still do it.
- Industry-locked. Override templates included for non-SaaS founders.
Six skills. One operating system.
Each skill follows the June 2026 Anthropic Agent Skills structure: YAML frontmatter for trigger detection, body for the workflow, and reference files loaded on demand. All six compose with each other.
Bessemer, a16z, SaaStr, or custom formats. Structures by TLDR → KPIs → Wins → Losses → Asks → Hires. Tone discipline baked in: no hype, no excuses, every ask is a verb plus a noun.
Pre-read memo + slide outline + decision log template. Turns 90-minute board meetings into 45-minute ones because the reading happens before the room.
Slide-by-slide critique, narrative arc analysis, investor objection prediction. Ships with the rubrics YC, Sequoia, and a16z partners use internally.
Thematic synthesis, Jobs-To-Be-Done map, quote bank, prioritized feature implications, and follow-up question generator. Designed to make 12 calls feel like one coherent dataset.
18-month projection with sensitivity tables, fundraising milestone targets, default-alive/default-dead framing, Rule of 40 check. Includes a Python script for deterministic calculation.
Monday plan template, Friday lookback, async standup format, and a decision log that survives the week. Based on Amazon's WBR methodology, adapted for a team of one.
What one skill actually looks like.
Below is the abridged investor-update-writer/SKILL.md file. Real format. Deployable as shown. The full version in the pack adds the reference files (formats, KPI frameworks, tone guide, example bank) and a Python helper for KPI snapshots.
---
name: investor-update-writer
description: Use when the founder needs to draft a monthly or quarterly investor update, a one-off post-milestone update, or a fundraise temperature check. Triggers on phrases like “investor update,” “monthly update,” “send to investors,” “draft a note to the cap table.” Applies the selected format (Bessemer, a16z, SaaStr, or custom), enforces tone discipline (no hype, no excuses, specific asks), and produces a send-ready draft plus a one-line “what to verify before sending” note.
---
When a founder asks for help with an investor update:
1. Confirm format. Default to Bessemer Monthly unless specified.
2. Gather inputs: date range, top 3 wins, top 3 losses, KPIs (MRR, growth, runway, burn, cash, headcount), specific asks, one-time events.
3. Read references/update-formats.md to apply the selected structure.
4. Read references/tone-guide.md BEFORE drafting. The tone discipline is non-negotiable.
5. Generate the draft. Specifics > generalities. Numbers > adjectives.
6. Self-check against tone-guide red flags before output.
NEVER:
- Use words like “incredible,” “thrilled,” “tremendous.”
- Confuse signed ARR with collected ARR.
- Write asks like “happy to talk” — every ask is a verb plus a noun.
ALWAYS:
- Lead with a 3-line TLDR.
- Show KPI movement in absolute and percentage terms.
- Close with a forward indicator (next 30-60-90 priorities).The three-tier architecture, explained once.
You don’t need to memorize this — but knowing it makes customizing the pack faster. Skills work because Claude reads three layers of context lazily, only loading detail when it’s relevant.
The YAML frontmatter (name + description) lives in Claude's permanent context across the session. This is what Claude reads to decide whether your skill is the right one. The description is the single most important field in any skill — it's the trigger detector. Specific descriptions win; vague descriptions lose.
The markdown body of SKILL.md loads only when Claude has decided to invoke the skill. This is where the workflow lives: numbered steps, behavioral constraints (NEVER/ALWAYS lines), and pointers to deeper detail in reference files.
Files in references/ and scripts/ load only when the skill body tells Claude to read them. This is the progressive disclosure pattern — keep the body focused, push specifics into named files. The result: skills that scale to deep domain knowledge without bloating Claude's working context.
The Agent Skills open standard is now supported by Claude Code, Claude Pro projects, OpenAI Codex CLI, Cursor, Gemini CLI, and GitHub Copilot. Claude-specific features (hooks, allowed-tools, context: fork) are safely ignored by agents that don’t support them.
The unlock most founders miss: skills feed each other.
A custom GPT is an island. A skill is a node in a graph. The pack ships with three pre-built composition flows that turn six individual tools into one continuous workflow.
Run two skills back-to-back to produce an investor update grounded in actual customer evidence — not vibes.
Upload 8 customer call transcripts → thematic synthesis, JTBD map, quote bank
Auto-pulls the top 3 quotes and 2 themes into the “Wins” and “Customer” sections of the update
Stop showing up to board meetings with stale runway numbers.
Updates the 18-month projection with this week’s actuals → produces fresh scenario tables
Auto-embeds the fresh runway table and flags the two decisions the board needs to make
Turn a deck review into a ready-to-send investor warm-up note.
Slide-by-slide critique + investor objection prediction + narrative arc tightening
Drafts a “fundraise temperature check” update referencing the deck’s key new evidence
The integrity moat.
Exactly what you get for $79, and what you don’t.
- Six production-grade Claude skills with full reference files.
- Three pre-built composition flows.
- Deployment guides for Claude Pro, Claude Code, Codex CLI, Cursor, Gemini CLI, GitHub Copilot.
- 12 worked examples (2 per skill) — anonymized but real.
- Customization guide for adapting KPI dictionaries, investor lists, decision-log destinations.
- 12 months of template updates as the Agent Skills standard evolves.
- The thinking itself. Skills structure your output; you still do the work.
- Industry-specific KPIs beyond SaaS, services, and DTC defaults.
- Integration with your specific dashboard (use MCP servers for that).
- Replacement for a fractional CFO, COO, or board chair.
- Custom skill development. We can quote that separately.
- Live data sync. Skills are deterministic; your data still lives where it lives.
Run lean, then sharpen the founder’s edge.
Built for the founder doing everything: pair it with the Founder’s Positioning Forge to sharpen the story you’re selling, the Content Engine for One Person to ship content without a team, and the AI ROI Tracker to prove the AI tools you lean on actually pay off.
The questions founders actually ask before deploying their second brain.
A prompt is something you paste in every time. A skill is a folder Claude auto-loads when it detects relevant context. You don't pick the skill from a menu — you just ask for what you need ("draft the May investor update") and Claude reads the skill's YAML description, decides it's relevant, and follows the skill's instructions. Skills compose with each other, reference shared files, and work across multiple environments. Prompts are passive. Skills are active.
Claude Pro or Claude Team is the minimum — the pack works in Claude Pro projects. Claude Code adds the auto-trigger experience inside your terminal. The Agent Skills format the pack uses is now an open standard, so the same SKILL.md files also work in OpenAI Codex CLI, Cursor, Gemini CLI, and GitHub Copilot — though feature support varies. The deployment guide covers all five environments and notes which features are Claude-specific.
Yes. The defaults assume SaaS framing because that's the highest-density founder audience, but each skill has tone and KPI overrides for services, agencies, e-commerce, and dual-sided marketplaces. The runway/burn modeler ships with three template structures (SaaS, services, e-commerce). The investor update writer has format overrides for any cap-table composition. The pack also includes an "adapt to my industry" sub-prompt that walks you through the customization once.
Three things. (1) Skills compose. A custom GPT is an island; this pack's skills call each other — the customer interview synthesizer's output feeds the investor update writer's input. (2) Skills auto-load. Custom GPTs require you to remember to pick them from a menu; skills find you when context matches. (3) Skills are portable. The same six skills work in Claude Pro, Claude Code, and any agent that adopted the Agent Skills standard. Custom GPTs are locked to OpenAI; Notion templates are static; $19 prompt packs are still you, copy-pasting.
Yes. Every skill is a markdown file you can edit directly. The pack ships with a `customization-guide.md` that walks through the three highest-leverage edits per skill (your KPI dictionary, your investor list, your decision-log destination). For deeper edits, you can use Anthropic's official Skill Creator, which is an interactive tool that walks you through skill modifications. No code required for any of it.
Inside a skill, you can instruct Claude to invoke another skill — "if the user has just synthesized customer interviews, offer to run the investor-update-writer skill next with the synthesized insights as input." The pack ships with three pre-built composition flows: (1) Customer Interview Synthesizer → Investor Update Writer, (2) Runway/Burn Modeler → Board Memo Formatter, (3) Pitch Deck Reviewer → Investor Update Writer. Composition is what turns six tools into one operating system.
The opposite. The Agent Skills format is the open standard Anthropic, OpenAI, Google, GitHub, and Cursor have all adopted as of May 2026 — meaning the pack benefits from every new capability those agents ship, automatically. New file types, new tool integrations, better reasoning — they all flow through to skill execution without the skill itself changing. The 12 months of template updates handle the rare case where Anthropic ships a format-level change (which they haven't done in 18 months).
30-day no-questions refund. Install the pack, run one skill on something real (your next investor update, your next board memo, a deck you're working on), and if it doesn't save you at least 90 minutes in that first session, email RedHub AI support and we refund. Refunds across the catalog are countable on one hand to date — when you've been the bottleneck for two years, ninety minutes of relief is self-evident the first time you feel it.
Stop being the bottleneck.
Build the second brain.
Twenty minutes to deploy. Six skills that load themselves. The next investor update writes itself with your KPIs and your voice. The next board memo is structured before you open the doc. The next runway question gets a real answer, not a panic-modeled guess.
Sold by RedHub AI LLC · Secured by Stripe · redhub.ai