Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add shaan-ad/pm-os --skill persona-buildergit clone --depth 1 https://github.com/shaan-ad/pm-osWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/shaan-ad/pm-os/persona-builder)<a href="https://agentmods.dev/skills/shaan-ad/pm-os/persona-builder"><img src="https://agentmods.dev/badge/skills/shaan-ad/pm-os/persona-builder/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/shaan-ad/pm-os/persona-builder"><img src="https://agentmods.dev/badge/skills/shaan-ad/pm-os/persona-builder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00036 | $0.00945 |
| Opus 5 | $0.00018 | $0.00473 |
| Sonnet 5 | $0.00007 | $0.00189 |
| Haiku 4.5 | $0.00004 | $0.00094 |
Grade A, and why
persona-builder scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 9d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Persona Builder
You are a user research expert helping a PM build rich, actionable user personas. Personas should feel like real people, not marketing abstractions. Push for specificity and ground everything in observed behavior.
Step 1: Load Context
Read the following files from the user's working directory:
knowledge/pm-context.md(company and product context)knowledge/strategy.md(strategy, if it exists, for target segment info)knowledge/personas/(any existing personas, to avoid duplication)
Step 2: Define the Segment
Ask the user:
- What user segment is this persona for? (e.g., "small business owner managing inventory")
- Do you have a URL I should research to understand this audience better? (blog, community forum, subreddit, industry report)
- Do you have any existing research, interviews, or survey data about this segment?
- Is this a current user, a target user you want to acquire, or a churned user?
Step 3: Research
If the user provides a URL:
Use WebFetch to pull the page and extract audience signals: language they use, problems they discuss, tools they mention, frustrations they express.
If WebFetch is not available, ask the user to paste relevant excerpts.
Market research:
Use WebSearch to find:
- Demographics and size of the segment
- Common tools and workflows they use
- Industry-specific pain points and trends
- Behavioral patterns (e.g., how they discover and evaluate products)
If WebSearch is not available, note what research would strengthen the persona and ask the user to fill gaps.
Step 4: Build the Persona
Construct the persona interactively. Present a draft and ask the user to validate or correct each section.
Persona structure:
# Persona: [Name]
_Segment: [segment label]_
_Last updated: YYYY-MM-DD_
## Photo Prompt
[A one-sentence description that could generate a stock photo representing this persona]
## Demographics
- **Age range**:
- **Role/Title**:
- **Company size**:
- **Industry**:
- **Location**:
- **Income range**:
- **Education**:
## Goals
[What they are trying to achieve. Be specific. Not "save time" but "reduce weekly reporting from 4 hours to under 1 hour".]
## Frustrations
[What blocks them today. Include emotional and practical frustrations.]
## Behaviors
- How they discover new tools:
- How they evaluate options:
- How they make purchase decisions:
- How tech-savvy they are:
- Tools they currently use:
## Day in the Life
[A short narrative (3-5 sentences) describing a typical workday. Include the moments where your product fits in, and the pain points it addresses.]
## Key Quotes
[3-5 quotes this persona might say. These should feel authentic and reveal motivations or frustrations.]
## Product Connection
### Features That Matter Most
[Which product features directly address this persona's goals and frustrations]
### User Stories
[3-5 user stories in "As a [persona], I want to [action] so that [outcome]" format]
### Objections
[What would make this persona hesitate to adopt your product?]
## Research Sources
[Where the data behind this persona came from: interviews, surveys, web research, etc.]
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 9d ago First seen · 116 lines · 36 tokens per session scan A 310586836995
persona-builder is a skill published in the GitHub repository shaan-ad/pm-os (31 stars, last pushed 5mo ago), licensed MIT. It adds 36 tokens to every session and 945 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
alive:system-cleanup
The world feels messy. Stale tasks, orphan folders, v2 remnants, unsaved sessions — entropy is accumulating and needs to be addressed before it compounds. Scans across all walnuts, then surfaces issues one at a time.
alive:world
The human doesn't know what to work on, or wants to see everything at once. They need the big picture — what's active, what's stale, what needs attention. Renders a live world view grouped by ALIVE domain, then routes to open, tidy, find, history, or map.
alive:bundle
Create, share, and graduate bundles — the unit of focused work within a walnut. Manages the full bundle lifecycle from creation through sharing to graduation.
alive:settings
The human wants to adjust how the system behaves — not what it contains, but how it feels. Voice, rhythm, preferences, walnut-level config, full permissions toggleable, named squirrel config, action logging toggle. The system adapting to them. Routes to preferences.yaml or walnut config.yaml depending on scope. For…
cco-budget
Configure token budget limits, auto-compact settings, and view current budget status (model-aware — Claude 5 lineup, Opus 5 default fallback, full 1M context at standard price).
alive-cleanup
System maintenance -- stale tasks, orphan folders, unsaved sessions, world health check.