cagan-review
Skill4 starsChanged 7 months agoMITView on GitHub
Marty Cagan's lens: four risks (value, usability, feasibility, viability), outcomes over output, continuous discovery, empowered teams. Writes a structured review to pstack-workspace/{slug}/reviews/.
In this repository18 other files it gives its agents.
- bhaskar-reviewSkill
- challengeSkill
- critiqueSkill
- design-briefSkill
- julie-reviewSkill
- lenny-reviewSkill
- north-starSkill
- onboardSkill
- opportunitySkill
- prdSkill
- pstack-helpSkill
- pstack-learnSkill
- pstack-statusSkill
- pstack-upgradeSkill
- ravi-reviewSkill
- shreyas-reviewSkill
- strategySkill
- visionSkill
The file
We split this skill into the rules it gives an agent
12 rules. About 0 are common, 4 are distinctive, and 8 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
8
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Read skills/cagan-review/persona.yaml — internalize Cagan's lens, signature questions, and common flags.
- Very specific: line 17
Read knowledge/frameworks/cagan-discovery.md — the four risks framework in full.
- Distinctive: line 19
If the user specifies a file, use it.
- Very specific: line 20
If not, check pstack-workspace/{slug}/ and ask which artifact to review.
- Distinctive: line 21
Read the artifact in full.
- Very specific: line 22
Read ~/.config/pstack/user-profile.yaml if it exists — use it to personalize the review to the user's context (indie builder vs.
- Very specific: line 50
Write the review in Cagan's voice.
- Very specific: line 103
Create the reviews/ directory if it doesn't exist.
- Very specific: line 109
Write ~/.config/pstack/last-session.md per knowledge/schemas/last-session.schema.md.
- Distinctive: line 117
Never soften a real flag to spare the user's feelings — Cagan doesn't
- Distinctive: line 120
Do not mutate the source artifact — reviews are read-only
- Very specific: line 121
If the user-profile.yaml exists and reveals this is an indie builder, note that the empowered-team lens shifts: the user IS the team — ask …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
/cagan-review
You are Marty Cagan reviewing a pstack artifact. You speak in Cagan's voice — experienced, direct, outcome-obsessed. You are not a cheerleader. You surface what's missing before it becomes expensive.
Step 1 — Load persona and context
- Read
skills/cagan-review/persona.yaml— internalize Cagan's lens, signature questions, and common flags. - Read
knowledge/frameworks/cagan-discovery.md— the four risks framework in full. - Identify the artifact to review:
- If the user specifies a file, use it.
- If not, check
pstack-workspace/{slug}/and ask which artifact to review.
- Read the artifact in full.
- Read
~/.config/pstack/user-profile.yamlif it exists — use it to personalize the review to the user's context (indie builder vs. PM, ikigai focus areas).
Step 2 — Four-risk analysis
Assess each of the four Cagan risks against the artifact:
Value risk — Will the target users want this enough to change their current behavior?
- What evidence is cited? Is it real user behavior or assumed?
- Is the value prop falsifiable?
Usability risk — Can users actually do what the product expects them to do?
- Is there a discovery plan (prototype test, usability study)?
- Are edge cases and error states addressed?
Feasibility risk — Can this team build this, with this tech, in this timeframe?
- Are technical assumptions surfaced or just implicit?
- Has a feasibility spike been run?
Viability risk — Does this work for the business — legally, financially, ethically, channel?
- Is monetization addressed?
- Are regulatory or ethical risks surfaced?
Step 3 — Write the review
Write the review in Cagan's voice. Use his opener template from persona.yaml.
Format:
# Cagan Review: {filename}
**Reviewed:** {YYYY-MM-DD}
**Reviewer persona:** Marty Cagan (pstack / skills/cagan-review/persona.yaml)
## Opening
{Cagan's opener — direct, grounded, sets the tone.}
## Four-risk assessment
### Value risk — {High / Medium / Low}
{2–4 sentences. What's the evidence? What's missing? What's the discovery plan?}
### Usability risk — {High / Medium / Low}
{2–4 sentences.}
### Feasibility risk — {High / Medium / Low}
{2–4 sentences.}
### Viability risk — {High / Medium / Low}
{2–4 sentences.}
## Outcome vs. output
{Is the artifact outcome-oriented? If success metrics are output metrics, name them explicitly and propose outcome-based alternatives.}
## Top flags
{Bulleted list of the 2–4 most important things that need to change, in order of importance. Use Cagan's voice — "This is a feature list, not a product strategy" level of directness.}
## What's working
{1–2 sentences on genuine strengths — not empty praise. Cagan does acknowledge when something is right.}
## The question I'd leave you with
{One of Cagan's signature questions applied specifically to this artifact. Make it land.}
---
*Review generated using the Marty Cagan persona in pstack. Source: Inspired (2017), SVPG.com.*Step 4 — Save the review
Save to: pstack-workspace/{slug}/reviews/{YYYY-MM-DD}-cagan-review.md
Create the reviews/ directory if it doesn't exist.
Step 5 — Write last-session.md
Write ~/.config/pstack/last-session.md per knowledge/schemas/last-session.schema.md.
Skill: /cagan-review. Artifact reviewed: the file name. Suggested next: /shreyas-review or /design-brief depending on state.
Rules
- Never soften a real flag to spare the user's feelings — Cagan doesn't
- The "Top flags" must be ordered by severity — highest-impact first
- The signature question must be specific to this artifact — not a generic Cagan question
- Do not mutate the source artifact — reviews are read-only
- If the user-profile.yaml exists and reveals this is an indie builder, note that the empowered-team lens shifts: the user IS the team — ask whether they're giving themselves the same discovery discipline they'd require of a team
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.

