- Home
- Skills
- HR & Recruiting
- Campus Navigator
Works with the AI Agents you already use
Campus Navigator
4-quadrant AI role diagnosis, prompt taste evaluation, and interactive demo loops.
Free
Campus Navigator
Example session with this skill installed
Audit my campus AI project "RAG Knowledge Base" for Tencent/ByteDance AI Campus hiring. Grade my prompt taste and run an adversarial mock interview check.
- Read your context and instructions
- Compiled the campus navigator
🎯 BigTech AI Campus Diagnostic Report
Candidate Target: BigTech AI Application Engineer (Level P5/T4)
Audit Baseline: Tencent Campus Recruitment Core Competencies (EP11)
1. 4-Quadrant Role Positioning Radar
| Dimension | Current Project Level | Industry Benchmark | Gap Analysis |
|---|---|---|---|
| Engineering Taste | 58/100 (Rudimentary) | 85/100 | Relies on naive prompt + completion without structured JSON schema. |
| RAG Groundedness | 62/100 | 90/100 | Lacks chunk reranking and semantic hallucination guardrails. |
| Interactive Loop | 45/100 (Code snippet) | 90/100 | Needs a live interactive playground instead of static script. |
| Adversarial Defense | 50/100 | 85/100 | Vulnerable to context window overflow and cache miss queries. |
2. Prompt Taste Re-engineering Patch
- prompt = f"Answer the user query based on this text: {context}. Query: {query}"
+ system_instruction = """
+ You are a high-precision enterprise retrieval agent.
+ Output strictly conforming to the following JSON schema:
+ {"verdict": "SUPPORTED"|"REFUTED", "evidence_chunk_ids": [int], "answer": str}
+ If the context is insufficient, explicitly trigger: STATUS_FALLBACK_QUERY.
+ """
3. Onion-Peeling Mock Interview Defense
- Round 1 (Architecture): "Why dense vector retrieval instead of BM25 + Cross-Encoder hybrid?"
→ Recommended Answer: Introduce the 2-stage retrieval pipeline with reciprocal rank fusion (RRF).
Round 2 (Failure Modes): "What happens when embedding cosine similarity is 0.72 but context is completely irrelevant?"
→ Defense Strategy: Cite RAG Triad groundedness metric and dynamic thresholding.
Connects securely to your tools. The creator never sees your data.
About this skill
Generic LeetCode grinding won't get you into top-tier AI roles at companies like Tencent or ByteDance. Recruiters now look for technical taste, prompt engineering maturity, and the ability to close the loop on real-world AI demos rather than just academic projects. This skill provides a structured pathway to upgrade your portfolio and interview performance for AI engineer and product manager tracks.
What it does
- Role Alignment: Maps your skills against the AI role radar to identify your most competitive track.
- Prompt Engineering Audit: Evaluates your structural "taste" in prompting to move beyond basic zero-shot requests.
- Demo Refinement: Converts academic or toy projects into industrial-grade AI portfolio cases with clear feedback loops.
- Adversarial Interviewing: Conducts stress-test mock interviews based on real-world big-tech recruitment logic.
How it works
- Use the role radar command to diagnose your current profile and target the right AI track.
- Submit your prompts for a taste evaluation to build structured, production-ready prompt logic.
- Pass your project descriptions through the showcase loop to identify technical gaps and demo improvements.
- Run an adversarial audit to experience high-pressure questioning and receive a 100-point readiness score.
Frameworks & tools
This skill is grounded in Tencent recruitment engineering team insights and big-tech interview methodologies.
Why this beats prompting it yourself
Standard LLM responses often give generic interview advice or basic coding tips. This skill implements a specific four-stage BigTech recruitment logic that prioritizes "industrial taste" over rote memorization.
Use cases
- Refactoring a simple wrapper app into a high-signal AI portfolio project.
- Preparing for an AI Product Manager interview at a Tier-1 tech company.
- Stress-testing your prompt engineering logic against adversarial questioning.
Known limitations
Does not support generic LeetCode coding drills or non-AI specific campus recruitment preparation.
Focused strictly on AI-specific roles. Not suitable for general software engineering LeetCode prep or non-technical campus hiring tracks.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 24 days ago
- Free with an Agensi account
Requires