pitch
Skill0 starsChanged 3 months agoMITView on GitHub
Lightweight pre-planning skill that turns a fuzzy idea into a structured Feature Pitch (pitch.md) before any deeper planning or PRD step runs. Asks 6 questions one at a time with adaptive multiple-choice anchors, then writes the pitch file. Use when the user says "/pitch", "I have an idea for", "let's pitch this", "new feature idea", "I want to build", or whenever a software idea needs to be shaped before a domain/spec review. NOT for full PRDs, abstract strategy, or ideas that already have a detailed spec.
14 other projects use a version of this.
Adopted by
Most recently taken up by
- serhiileniv/skillstheir own version, first seen Nov 8, 2026
- serhiileniv/ccstheir own version, first seen Nov 8, 2026
- eggs-gd/career-spacetheir own version, first seen Nov 8, 2026
- aya-taleb/biz-ready-skillstheir own version, first seen Nov 1, 2026
- bengialb/pitch-skilltheir own version, first seen Oct 25, 2026
- FakharEAli/book-skillstheir own version, first seen Oct 25, 2026
Used with
Projects with it also use launch and spec.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
MicrosoftSkilllaunch
Launch and automate VS Code Insiders with the Copilot Chat extension using @playwright/cli via Chrome DevTools Protocol. Use when you need to interact with the VS Code UI, automate the chat panel, test the extension UI, or take screenshots. Triggers include 'automate VS Code', 'interact with chat', 'test the UI', 'take a screenshot', 'launch with debugging'.
Also in 3 projects that use this
SentrySkillspec
Produce a verified spec — problem, desired outcome, and acceptance criteria — before building. Use when given a GitHub/Linear issue link, when starting from a fuzzy in-conversation idea, or asked to "figure out what to do" / "scope this" / "what should I build" — whenever the *what* isn't yet pinned down. Grills the ambiguities, checks the premise against the code, and produces acceptance criteria you sign off on — then hands to design-first.
Also in 3 projects that use this
Other versions14 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 14
- lglucas/ai-dev-operating-system★ 12same name
- eggs-gd/career-space★ 4same name
- FakharEAli/book-skills★ 3same name
- aya-taleb/biz-ready-skills★ 1same name
- madeehahassan/plsbuild-os★ 1same name
- ORBWEVA/founder-pitch★ 1same name
- T4LEL/Claude-Arsenal★ 1same name
- vertile-ai/skills★ 1same name
- bengialb/pitch-skill★ 0same name
- cyb3rcricket/cricket-mode★ 0same name
4 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository7 other files it gives its agents.
- auto-waveSkill
- check-domainSkill
- handoverSkill
- pmSkill
- process-interviewerSkill
- seo-geo-auditSkill
- usage-dashboardSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
18 rules. 0 are common, 14 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 14 edited versions we read keep these.
14
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
Before the first question:
- Very specific: line 21
Read CONTEXT.md in the current project (or CONTEXT-MAP.md if present): it supplies the actors for Q2 and the out-of-scope suggestions for …
- Very specific: line 22
Read docs/adr/ if present: it supplies no-go candidates for Q6.
- Distinctive: line 23
If no project context is detectable: use the defaults (see references/defaults.md).
- Distinctive: line 31
No multiple choice: free text.
- Distinctive: line 32
If the answer is too abstract ("users are confused"): dig in with "Show me a concrete situation: who's doing what, and what goes wrong?"
- Distinctive: line 37
If no CONTEXT.md: End user / Admin / Internal team / Multiple.
- Distinctive: line 38
Always: "+ Other (free text)" as the last option.
- Distinctive: line 42
No MC: free text, with a starter: "Imagine you're explaining it to someone at a whiteboard in 30 seconds."
- Distinctive: line 43
If the answer is too precise (user stories, DB schema): stop and say "That's already PRD-level: save it for the planning step.
- Distinctive: line 50
Undecided: not settled yet
- Distinctive: line 55
Always include a free-text option.
- Distinctive: line 61
If no ADRs: compliance notes (GDPR, etc.), hard technical limits.
- Very specific: line 70
If docs/pitches/ doesn't exist: create it lazily.
- Distinctive: line 71
If no project root is detectable: the current directory.
- Distinctive: line 94
User already has a pitch.md: ask whether to update it or start a new version.
- Distinctive: line 96
No CONTEXT.md present: use defaults for Q2 + Q5 (see references/defaults.md).
- Very specific: line 98
User skips a question: ask once more. If still no input: leave the section empty and mark it in pitch.md as _Not answered: resolve during …
“Common” is measured: kept by at least 6 in 10 of the 14 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
What this skill does
Asks 6 questions: one at a time, with multiple-choice anchors drawn from the project context, and then writes a pitch.md. That file is the input for your next planning step (a domain/spec review, a PRD, or a companion skill like grill-with-docs if you use one).
Why not jump straight to planning? A planning/spec step can only challenge something once there's something to challenge. Without a pitch, the first half of the session goes to figuring out what should even be built. The pitch separates "what do we want" from "does it fit our domain model".
Flow
Step 1: Load context
Before the first question:
- Read
CONTEXT.mdin the current project (orCONTEXT-MAP.mdif present): it supplies the actors for Q2 and the out-of-scope suggestions for Q5. - Read
docs/adr/if present: it supplies no-go candidates for Q6. - If no project context is detectable: use the defaults (see references/defaults.md).
Step 2: Ask the 6 questions
Always: one question per message. Name a recommendation. Confirm before moving on.
Q1: The problem
"What's the concrete problem? Describe a situation, not a feature."
- No multiple choice: free text.
- If the answer is too abstract ("users are confused"): dig in with "Show me a concrete situation: who's doing what, and what goes wrong?"
Q2: Audience
"Who is this for, who hits this problem?"
- Generate MC options from the actors in CONTEXT.md.
- If no CONTEXT.md: End user / Admin / Internal team / Multiple.
- Always: "+ Other (free text)" as the last option.
Q3: Rough solution
"What's your rough idea for the solution? 2–3 sentences, no wireframe."
- No MC: free text, with a starter: "Imagine you're explaining it to someone at a whiteboard in 30 seconds."
- If the answer is too precise (user stories, DB schema): stop and say "That's already PRD-level: save it for the planning step. Rougher, please."
Q4: Scope of the first version
"How built-out should the first version be?"
- MC:
- MVP: fastest version that solves the problem, technical debt knowingly accepted
- Polished: production-ready, all edge cases
- Undecided: not settled yet
Q5: Out of scope
"What's explicitly NOT part of this: things that might seem adjacent but aren't coming yet?"
- MC suggestions from CONTEXT.md: things marked there as "out of scope" or "phase X".
- Always include a free-text option.
- Recommendation: "I'd suggest [X from CONTEXT.md] as out of scope, because the phase plan puts it in phase Y."
Q6: No-goes
"What's an absolute no-go, something that must under no circumstances be part of this solution?"
- MC candidates from ADRs (rejected alternatives) + known constraints.
- If no ADRs: compliance notes (GDPR, etc.), hard technical limits.
- Always include a free-text option.
- May be left empty, then drop the section from pitch.md.
Step 3: Write pitch.md
Right after Q6: no further questions.
Location: docs/pitches/pitch-YYYY-MM-DD-<slug>.md in the project root.
If docs/pitches/ doesn't exist: create it lazily.
If no project root is detectable: the current directory.
Slug = 3–4 words from the problem, kebab-case.
Format: see references/pitch-template.md
Step 4: Handoff
After writing:
pitch.md is at docs/pitches/pitch-YYYY-MM-DD-<slug>.md.
Next step: hand it to your planning/spec step (a domain review, a PRD, or the
grill-with-docs skill if you use one), point it at that file.Edge cases
User runs /pitch with text: e.g. /pitch I want SMS reminders for bookings
→ Take that text as pre-context for Q1, but still ask Q1 and confirm.
User already has a pitch.md: ask whether to update it or start a new version.
No CONTEXT.md present: use defaults for Q2 + Q5 (see references/defaults.md). Note in the output: "No CONTEXT.md found: actors and out-of-scope suggestions are defaults."
User skips a question: ask once more. If still no input: leave the section empty and mark it in pitch.md as _Not answered: resolve during planning_.
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.

