start — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited start (Agent Skill) and scored it 96/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Orient the user and guide them through first-time setup. Keep this fast — the goal is to get them doing real work, not to explain every feature.
Greet the user with a short plain-English explanation of what bettersense is, then ask one question to branch the flow.
bettersense is a set of skills and agents for technical leaders — AI PMs, engineering managers, TPMs, and senior ICs. It covers things like writing specs, preparing for 1:1s, giving feedback, logging wins, reflecting on your stakeholders, and coaching your reports.
>
Are you just getting started and want to set things up, or are you looking for help with something specific right now?
Do not list all 40+ skills upfront. Do not ask multiple questions. One question, two paths.
If the user is new or wants to set up:
The first thing worth doing is setting up your profile — a short "who you are" file that most skills read automatically. It means you don't re-explain your role and context every session.
>
It takes about 5 minutes. Want to do that now? If yes, just say "yes" and I'll hand you to the profile skill. You can say 'stop' at any time.
If they agree, invoke user-profile. Wait for it to complete (or for the user to exit early) before continuing.
Now that the profile says what they do, hand them a starter kit: the five skills for their role, with explicit permission to ignore everything else. This is the antidote to forty-skill overwhelm — confidence comes from five things that fit, not a catalog.
Frame it like this (adapting to their actual role):
bettersense has a lot in it, but you only need five things to start. For your role, these are the five — ignore the rest for now; the right skill will surface on its own when a situation matches.
| Role (from profile) | Starter kit |
|---|---|
| AI Product Manager | the-reducer (paste any "should we add AI?" ask into it) · ai-pm-frameworks · decision-log · metrics-design · weekly |
| Engineering Manager | one-on-one-prep · feedback-frameworks · stakeholder-reflect · wins-log · weekly |
| TPM / Program Manager | the-program-manager · influence-without-authority · the-status-crafter · stakeholder-reflect · weekly |
| Senior IC | wins-log · glue-audit · influence-without-authority · the-rfc-reviewer · weekly |
If the role straddles two personas (common), blend: pick the people-side trio from one and the craft pair from the other. If the role matches none, build a five-skill kit from what the profile says they actually spend time on.
Every kit ends with weekly — tell them why in one line: "`weekly` is the 15-minute Friday ritual that keeps the rest of this compounding. It's the one habit worth forming."
Next — do you have a manager, report, peer, or team you want to reflect on regularly? If so, register one stakeholder now and you'll have a live example of how the reflection system works.
If they agree, invoke stakeholder-register. Optional — don't push if they decline.
Once setup is done (even partially), close with two concrete things to try, drawn from their starter kit:
You're set up. Two things to try this week:
>
1. [First kit skill applied to something they mentioned — e.g. "Let me reflect on [name]" or "Paste that AI feature request and I'll run it past the-reducer."] 2. Book a 15-minute block Friday and run `/bettersense:weekly` — that's the whole maintenance habit.
>
Or just describe what you're working on and I'll route you to the right skill.
If the user already knows what they want (or just wants to explore):
Ask them what they're working on. Don't give a full skill list — route based on their answer. Common entry points:
| What they say | Skill to invoke |
|---|---|
| Writing a spec, PRD, one-pager | the-spec-writer |
| Preparing for a 1:1 | one-on-one-prep |
| Giving feedback to someone | feedback-frameworks |
| Coaching a report | coaching-mode |
| Logging something I shipped | wins-log |
| Thinking through a decision | decision-log |
| Reflecting on my manager / report / peer | stakeholder-reflect |
| Reviewing a prompt or system prompt | the-prompt-critic |
| Understanding my team's health | team-diagnosis |
If nothing matches, say: "Describe what you're trying to do and I'll find the right skill — or browse the full list with `/bettersense:` and tab-complete."
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.