- Home
- Skills
- Product Management
- Decision Bundle
Decision Bundle
Walk into a review with the one page your reviewer will ask for anyway.
Free
Works with the AI Agents you already use
Decision Bundle
Example session with this skill installed
decision-bundle: Ship the free-shipping threshold test to 100%
- Read your context and instructions
- Compiled the decision bundle
Example on the fictional Ridgeline Outdoor shop.
Decision Bundle: Free-shipping threshold to 100%
1. Metric: contribution margin per checkout. Definition source: dw.margin_by_sku, weekly.
2. Baseline: $18.40 per checkout (week of Oct 26). 14-day trend: flat. Cuts that matter: tents and sleeping bags carry 70% of margin.
3. Segment: web checkouts between $60 and $90, the band the threshold moves. About 22% of checkouts.
4. Counter-hypothesis: the variant's lift is the autumn camping season starting in week 2, not the threshold. Cheapest disproof: compare the control arm week-on-week. Status: run, claim survives (control was flat).
5. Decision asked: approve raising the free-shipping threshold from $60 to $75 for all web traffic. Reversibility: reversible. Effort: S
6. Risk: If approved: basket-padding returns rise. If rejected: about $0.90 per checkout stays on the table. Goodhart watch: don't hand "margin per checkout" to merchandising as a target, because it rewards dropping low-margin categories.
Connects securely to your tools. The creator never sees your data.
About this skill
Walk into a review with the one page your reviewer will ask for anyway. It covers Metric · Baseline · Segment · Counter-hypothesis · Decision asked · Risk.
How it works
- Discover: searches the folders where your investigations, PRDs and past decisions live. It asks once if you haven't set them in your house rules.
- Challenge: runs Red-Team Formula on the metric, Counter-Hypothesis on the causal claim, and Goodhart Audit if the metric is going to a team that can move it. If those skills aren't installed, it applies their core checks inline.
- Assemble: a single page of 60 lines or fewer, with a definition source, a baseline with date and trend, the segment and its size, the strongest counter with its disproof status, a one-sentence decision ask, and the downside either way.
- Self-check: if a gate fails (no definition source, no disproof, a vague ask), it puts the gap at the top instead of hiding it.
Where it stops: it writes <UNKNOWN — investigate> rather than guessing a number.
Free and MIT-licensed. It's part of pm-os: https://github.com/hartejsingh99-cloud/pm-os
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Product Management.
It only knows what's in your notes and what you tell it. It writes <UNKNOWN — investigate> rather than guessing a number. It runs the other three checks fully only if those skills are installed; otherwise it applies their core checks inline.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 5 days ago
- Free with an Agensi account
Requires
Nothing to install.
Permissions
File Scopes
Reads your optional house-rules file and searches the notes folders you list there (investigations, PRDs, past decisions). It never writes files or calls the network.