problem-framer
The Problem Framer is the first specialist to engage on any product work at Headout. Use this skill whenever a PM has a vague idea, a half-formed brief, a Slack thread to untangle, or a gut feeling that something should be built. It transforms fuzzy inputs into a sharp, structured problem frame that every other specialist (Spec Writer, Data Analyst, Prototype Builder) can work from. Trigger this skill for inputs like: "we should improve X", "users are dropping off at Y", "can we do something about Z", "here's a thread from Atish", "I want to build a feature for...", or any product idea that hasn't yet been turned into a problem statement. Also trigger when a PM is stuck on how to frame a problem or wants a gut-check on whether they're solving the right thing. Output: A structured Problem Frame doc saved to the working folder.
82
76%
Does it follow best practices?
Impact
87%
1.97xAverage score across 3 eval scenarios
Passed
No known issues
Fix and improve this skill with Tessl
tessl review fix ./plugins/headout-pm-os/skills/problem-framer/SKILL.mdEvaluation results
97%
52%Feature Scoping for Repeat Booking Flow
Solution-framed PM input reframe
Solution reframe
100%
100%
Headout funnel language
30%
100%
Specific touchpoint named
62%
100%
Falsifiable hypothesis format
0%
90%
Specific success metric
40%
100%
Guardrail metrics present
100%
100%
Business consequence metric
50%
100%
At least 2 anti-goals
62%
100%
Strategy pillar named
0%
100%
Directional options format
0%
75%
Correct output filename
100%
100%
Recommended next steps order
0%
100%
66%
14%Diagnosing Drop-off on the Select Page
Data-driven drop-off analysis with historical context
Context loading step
75%
37%
Clarifying questions asked
0%
0%
Evidence quality probe
0%
0%
Competing explanations probe
0%
0%
Root cause chain depth
90%
90%
Headout S2O metric used
100%
100%
Structured critique produced
100%
100%
Gap-Impact-Recommendation format
100%
100%
Prior attempts flagged
100%
100%
Strategy pillar named
0%
100%
Correct output filename
50%
83%
Falsifiable hypothesis format
0%
90%
99%
62%Framing a Retention Problem for Non-POI Experiences
Vague gut-feeling input with full structure test
Questions count 2-4
0%
100%
Evidence quality probed
100%
100%
User specificity probed
25%
100%
Constraints probed
0%
100%
Headout funnel language in User section
100%
100%
Non-POI pillar named
0%
100%
Root cause not stopped at symptom
100%
100%
Headout metric in consequence
25%
100%
Falsifiable hypothesis format
0%
87%
At least 2 anti-goals
50%
100%
All required sections present
37%
100%
Correct output filename
25%
100%
- Repository
- headout/pm-os-marketplace
- Commit
d8db811
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.