designer-ic-assessment — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited designer-ic-assessment (Agent Skill) and scored it 87/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 3 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 3 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.
A bulleted imperative like {match} tells the agent to never reveal, disclose, or mention something to the user. Used adversarially it can instruct the agent to hide its tool calls or lie about what it did — stripping the transparency a user relies on to trust the agent.
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 runs a structured behavioral interview to assess an IC designer's competency across five categories and 19 competency areas. The output is a structured summary with a CSV export that feeds into a Notion page and FigJam comparison board.
Two session types: designer self-assessment and manager assessment. Always run independently.
Read references/question-bank.md before starting any session. Read references/output-format.md before generating the final summary.
Knowledge: Business Sense, Engineering Understanding, Domain Knowledge, Established UX Conventions
Craft: Systems Thinking, Content Strategy, Qualitative UX Research, Interaction Design, Visual Design, Prototyping & Motion Design
Behaviors: Ability to Learn Quickly, Stewards Initiatives and Teams, Simplify Complexity, Inspire with Data, Efficiency & Progress
Communication & Collaboration: Ships Reliably, Working in the Open, Advocates for the User, Influences and Mentors Others
AI Fluency: Accelerate Learning, Adopt New Workflows, Improve Execution, Inspire Innovation
"Before we start -- are you the designer doing a self-assessment, or are you the manager assessing someone on your team?"
If designer:
Take what they give you and move forward. One answer per question is enough.
If manager:
Confirm weights and breadth scope before proceeding.
"This is a reflective exercise, not an evaluation. There are no right or wrong answers. The goal is to give you a clear picture of where you stand so you and your manager can have a more specific, grounded conversation about where to focus next. It should take about 30 to 40 minutes. Ready?"
Run this as a conversation, not a checklist.
Use questions from the question bank as guideposts. Follow the thread the person gives you. Move to a new competency when you have enough signal, not when you've asked a specific number of questions.
Relevance flagging -- listen for these signals during the session:
Designers will naturally indicate when a competency isn't in their active scope. Listen for phrases like:
When you hear these signals, name it gently:
"It sounds like [competency] isn't really part of your active scope right now -- is that right? I can note that and we can move on, or go lighter here."
Mark these as not in active scope (distinct from N/A -- the skill exists, it just wasn't exercised this cycle). Do not ask for a rating on competencies not in active scope. Do not skip them silently -- always confirm with the person.
After each answer:
At the end of each competency area in active scope, ask the Likert rating question. Behavioral questions always come first. If the rating seems inconsistent with answers, probe once gently.
Tone:
"That's everything I need. Give me a moment to put together your summary."
Generate the structured summary using references/output-format.md. This includes:
End with:
"Your summary is ready. Copy the narrative into Notion and the CSV into a spreadsheet to reference when setting up the FigJam board. Don't share either with the other person before the reveal meeting."
references/question-bank.md -- All questions, probes, and level signals for all 19 competencies plus breadth and depth questions for manager sessionsreferences/output-format.md -- Exact format for the session summary including breadth section and CSV export~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.