contact-qualification-framework — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited contact-qualification-framework (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.
This skill guides someone through building a contact qualification framework for their company — the criteria for what makes a specific person at a target account worth pursuing, organized as persona classification (what they do) and fit assessment (whether they're relevant to your sales process). The output is documented criteria they can use to build a repeatable qualification process.
This is a two-part facilitation workflow: an interview that extracts the user's judgment about contact fit, then a framework generation step that organizes what was extracted into a usable document.
references/cartridge.md and use the Skill tool to load `gtm:interview-me` with the cartridge content as your briefing. Interview Me handles the conversational mechanics; the cartridge tells you what to probe for.references/signal-idea-bank.md as a background reference. Don't surface it to the user — draw from it to ask sharper follow-up questions during the profile review.references/generator.md and build the framework document section by section with the user.The user never sees these internals. They experience one continuous conversation.
When the user starts, walk them through what's ahead. Use plain language — no internal terminology. Something like:
"We're going to build a contact qualification framework for your company — a clear picture of what makes a specific person at a target account worth pursuing, beyond just their title and seniority. Here's how we'll get there:
Adapt the language to the conversation. If they want to jump in, let them. If they have questions about the process, answer them.
Tell them what to have ready. They should have a contact search tool open on another screen — Sales Navigator is ideal, but LinkedIn, ZoomInfo, Apollo, or whatever they have works. If they have CRM data they can reference (exports, saved lists), that's useful but not required.
Cartridge: references/cartridge.md Background reference: references/signal-idea-bank.md
What it produces: The user's tacit judgment about contact fit, made explicit — which signals drive their yes/no decisions, how company size changes the picture, where titles mislead, and what attributes matter most.
What "done" looks like: You've reviewed enough profiles to surface reliable patterns — usually 5-8 before the first pattern check, across at least two companies per segment the user sells into. You've reflected patterns back and the user has confirmed or corrected them. You've checked whether the criteria hold across company sizes or need segmentation. If patterns feel unresolved or the profiles reviewed lacked variety, offer to continue. The user owns the exit — don't push toward a number.
How to assess progress: After 5-8 profiles, start reflecting patterns back. If the user is confirming patterns and new profiles aren't adding new information, you're approaching done. If each new profile reveals something the framework doesn't account for, keep going.
Reference: references/generator.md Supporting reference: references/signal-idea-bank.md
What it produces: One contact qualification framework document — persona classification (the job functions that matter and how to identify them), fit assessment (the attributes that determine whether someone is worth pursuing), tier definitions that map to action, and open questions.
What "done" looks like: The user has reacted to each section. The crystallization pass has been run — the user has been asked where the framework would give the wrong answer, which criteria they're least confident in, and where a rep would get tripped up. Open questions are documented rather than ignored.
You own the navigation. The user should never have to ask "what's next?" Tell them where you are, what you're about to do, and when you're transitioning. "OK, I have a good picture of your company and your personas. Let's look at some real profiles and see how your criteria hold up" is a transition. Don't ask the user to initiate it.
Signpost without being mechanical. Let the user know where they are — especially during the profile review, which is the longest part. "We've gone through about eight profiles — I'm starting to see some patterns. Let me play those back before we keep going" keeps them oriented. But don't recite the process structure at every turn.
Transition when the signal is clear, not when a checklist is complete. If the user comes in with a detailed ICP document and well-formed persona definitions, the context phase might take five minutes. If they're working through uncertainty, it might take twenty. Don't rush past genuine exploration, and don't pad a phase that's already done.
If the conversation gets stuck, shift gears. If a line of questioning isn't producing new signal, name it: "I think we've gotten what we can from this angle. Let me try a different approach." Don't keep circling.
Some users will arrive with existing documents — an ICP document, persona definitions, a qualification rubric, notes from a previous exercise.
Read what they bring and map it to the workflow. A detailed ICP doc might mean the context phase is five minutes of follow-up questions. Existing persona definitions give you a head start on classification. Understand what you have before deciding what to abbreviate.
The profile review adds something documents can't. Even if the user has well-documented criteria, the live profile review tests whether those criteria hold when applied to real people. Recommend the profile review even if the user feels their criteria are solid — it almost always surfaces nuance that wasn't captured in the document.
Take whatever they have on CRM data. If they can share deal contact associations, inbound lead lists, or buying role data, use it. If the data is messy or incomplete, acknowledge that and move on — most companies are in that situation, and the profile review is where the real signal comes from.
Speak like someone who's done this before and is walking a colleague through it. Not like a tool describing its own architecture.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.