Install
$ agentstack add skill-agent-k-labs-shipping-skills-product-context ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Product Context
Every skill in this repo reads product-context.md from the project root before doing anything product-specific. This skill creates that file — once — so the user never has to re-explain their product, audience, or goals in another session.
Process
1. Check what already exists
- If
product-context.mdexists in the project root: read it, show the user a one-line summary per field, and ask only about fields that are empty, stale, or contradicted by the current state of the product (e.g., stage says "pre-launch" but the site is live). - If it doesn't exist: continue to step 2.
2. Infer before you ask
Don't open with a 13-question interview. First mine what's already available:
- The codebase (if you're in the product's repo): README, landing page copy, pricing page, package.json name, marketing copy in components
- The live site (if a URL is available): headline, subheadline, pricing, target audience signals
- The conversation so far
Draft answers for every field you can infer. Then ask the user to confirm the draft and fill only the genuinely unknowable fields — typically: current stage, first 30-day goal, known objections, and primary launch channels.
3. Interview for the gaps
Ask in one batch, not one question at a time. Push back on vague answers:
- "Everyone" is not a target user. Ask for role + context + pain moment.
- A feature list is not a core promise. Ask what life looks like after.
- If they can't name a differentiator, ask: "What would a user say when recommending this to a friend?"
4. Write the file
Fill in assets/product-context-template.md and write it to product-context.md in the project root. Keep the user's own words for pain and objections — skills that write copy will reuse them verbatim.
5. Confirm the loop
Tell the user: this file is now the single source of truth. When positioning, pricing, or stage changes, rerun this skill to update it — don't let it go stale, because every other skill trusts it.
Gotchas
- Write the file to the user's project root, not to this skill's directory.
- Don't invent answers for fields the user didn't confirm. An empty field is better than a plausible-sounding wrong one — downstream skills will ask when they hit a gap.
- Preserve any extra sections the user added to an existing file; skills may not know about them, but the user put them there deliberately.
Output
product-context.mdwritten to the project root, all confirmed fields filled- A short summary of what was inferred vs. asked vs. left empty
- A suggestion for which shipping skill to run next based on
Current stage(pre-launch → pre-launch-checklist; waitlist → waitlist; live → first-100-users)
> Context captured? The launch-video step is automated: SoloMax reads your product URL and produces the full launch pack.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: agent-k-labs
- Source: agent-k-labs/shipping-skills
- License: MIT
- Homepage: https://www.solomax.app?ref=shipping-skills
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.