Build Skill-Augmented AI Agents with SkillNet for Search, Evaluation, Graph Analysis, and Task Planning

SkillNet, a framework for discovering, installing, evaluating, and organizing reusable AI skills, has been released as an open-source tool that enables developers to build skill-augmented AI agents. The platform supports keyword and semantic search for finding skills, quality evaluation across multiple dimensions, graph-based visualization of skill relationships, and automated task planning that breaks complex goals into subtasks with relevant skill pipelines. The implementation provides both an SDK and REST API fallback, allowing developers to integrate curated skills from GitHub into agent workflows for search, evaluation, graph analysis, and task planning applications.

In this tutorial, we implement a SkillNet https://github.com/zjunlp/SkillNet use case as a practical framework for discovering, installing, inspecting, evaluating, and organizing reusable AI skills. We start by setting up a robust SkillNet client with SDK and REST fallback support, then compare keyword search with semantic search to understand how skills can be found for different task requirements. From there, we install curated skills from GitHub, inspect their metadata, apply a quality gate across key evaluation dimensions, and visualize relationships between skills as a graph. Finally, we build a skill-augmented agent planner that breaks a complex goal into subtasks, discovers relevant skills, filters them, and assembles an execution pipeline. python import sys, subprocess def pip pkgs : subprocess.run sys.executable, "-m", "pip", "install", "-q", pkgs , check=False print "Installing dependencies skillnet-ai, networkx, matplotlib, requests ..." pip "skillnet-ai", "networkx", "matplotlib", "requests" import os, re, json, textwrap, pathlib, traceback import requests API KEY = os.environ.get "API KEY", "" BASE URL = os.environ.get "BASE URL", "https://api.openai.com/v1" MODEL = os.environ.get "SKILLNET MODEL", "gpt-4o" GITHUB TOKEN = os.environ.get "GITHUB TOKEN", "" GITHUB MIRROR = os.environ.get "GITHUB MIRROR", "" if not API KEY: try: from google.colab import userdata API KEY = userdata.get "API KEY" or "" except Exception: pass REST BASE = "http://api-skillnet.openkg.cn/v1" WORKDIR = pathlib.Path "./skillnet demo" ; WORKDIR.mkdir exist ok=True SKILLS DIR = WORKDIR / "my skills"; SKILLS DIR.mkdir exist ok=True def banner title : line = "=" 78 print f"\n{line}\n {title}\n{line}" We install the required dependencies and prepare the basic environment for the SkillNet tutorial. We configure API keys, model settings, GitHub options, and working directories to ensure the rest of the workflow runs smoothly. We also define a reusable banner function to keep the tutorial output organized and readable. banner "1 Initialize SkillNet client SDK with REST fallback " USE SDK = False client = None try: from skillnet ai import SkillNetClient client = SkillNetClient api key=API KEY or None, base url=BASE URL, github token=GITHUB TOKEN or None, USE SDK = True print "SDK loaded: skillnet ai.SkillNetClient" except Exception as e: print f"SDK unavailable {e r} ; using REST fallback for search/download." def norm item : if isinstance item, dict : g = item.get else: g = lambda k, d=None: getattr item, k, d return { "skill name": g "skill name" or g "name" or "?", "skill description": g "skill description" or g "description" or "", "author": g "author" or "", "stars": g "stars" or 0, "skill url": g "skill url" or g "url" or "", "category": g "category" or "", } def search q, mode="keyword", limit=5, min stars=0, sort by="stars", threshold=0.8 : if USE SDK: try: kw = dict q=q, limit=limit, mode=mode if mode == "keyword": kw.update min stars=min stars, sort by=sort by else: kw.update threshold=threshold res = client.search kw return norm x for x in res or except Exception as e: print f" SDK search failed - REST {e r}" params = {"q": q, "mode": mode, "limit": limit} if mode == "keyword": params.update min stars=min stars, sort by=sort by else: params.update threshold=threshold try: r = requests.get f"{REST BASE}/search", params=params, timeout=30 r.raise for status return norm x for x in r.json .get "data", except Exception as e: print f" REST search failed {e r}" return def show results results, title="" : if title: print f"\n-- {title} --" if not results: print " no results / endpoint unreachable " return for i, s in enumerate results, 1 : desc = textwrap.shorten s "skill description" , 70, placeholder="..." print f" {i}. {s 'skill name' :<34} ⭐{s 'stars' :<5} {s 'category' } " if desc: print f" {desc}" banner "2 Search: keyword vs. semantic vector " kw hits = search "pdf", mode="keyword", limit=5, sort by="stars" show results kw hits, "keyword: 'pdf' sorted by stars " vec hits = search "analyze financial reports from documents", mode="vector", limit=5, threshold=0.80 show results vec hits, "vector: 'analyze financial reports from documents'" We initialize the SkillNet client and provide a REST fallback, so the tutorial remains usable even if the SDK does not work. We define helper functions to normalize search results and perform both keyword and semantic searches. We then compare a keyword search for PDF-related skills with a vector search for analyzing financial reports from documents. banner "3 Install skills download from GitHub into ./skillnet demo/my skills " CURATED = "https://github.com/anthropics/skills/tree/main/skills/skill-creator", "https://github.com/anthropics/skills/tree/main/skills/algorithmic-art", for s in kw hits + vec hits : if s "skill url" and s "skill url" not in CURATED: CURATED.append s "skill url" CURATED = CURATED :4 def download url, target dir : if USE SDK: try: kw = {} if GITHUB MIRROR: kw "mirror" = GITHUB MIRROR return client.download url=url, target dir=str target dir , kw except TypeError: return client.download url=url, target dir=str target dir except Exception as e: print f" download failed for {url}: {e r}" return None print " SDK not present — skipping live download for this URL " return None installed = for url in CURATED: print f" downloading: {url}" path = download url, SKILLS DIR if path: installed.append path print f" - {path}" print f"\nInstalled {len installed } skill s ." We create a curated list of useful SkillNet-compatible skills and expand it using the search results collected earlier. We download selected skills from GitHub into a local skills directory when the SDK is available. We keep the installation process small and quick, so the tutorial remains practical for Google Colab. python banner "4 Inspect installed skills SKILL.md frontmatter " def parse skill md skill path : p = pathlib.Path skill path md = None if p.is dir : for cand in p.rglob "SKILL.md" : md = cand; break elif p.name.upper == "SKILL.MD": md = p if not md or not md.exists : return {"path": str skill path , "name": p.name, "meta": {}, "found": False} text = md.read text encoding="utf-8", errors="ignore" meta = {} m = re.match r"^---\s \n . ? \n---", text, re.DOTALL if m: for line in m.group 1 .splitlines : if ":" in line: k, v = line.split ":", 1 meta k.strip = v.strip .strip '"' .strip "'" return {"path": str md , "name": meta.get "name", p.name , "meta": meta, "found": True} inspected = parse skill md pp for pp in installed if installed else for info in inspected: print f" • {info 'name' } {'SKILL.md found' if info 'found' else 'no SKILL.md'} " desc = info "meta" .get "description", "" if desc: print f" {textwrap.shorten desc, 90, placeholder='...' }" if not inspected: print " nothing installed locally — likely no SDK/network; sections 2 & 7 still run " We inspect the installed skills by searching for their SKILL.md files and reading their metadata. We parse the front matter to extract useful information, such as the skill name and description. We then print a clean summary of each installed skill to understand what has been added locally. banner "5 Evaluate skills on 5 quality dimensions quality gate " DIMS = "safety", "completeness", "executability", "maintainability", "cost awareness" LEVEL SCORE = {"Excellent": 4, "Good": 3, "Fair": 2, "Poor": 1, "Bad": 0} def evaluate target : if USE SDK and API KEY: try: return client.evaluate target=target except Exception as e: print f" evaluate failed for {target}: {e r}" return None def mock eval name : import hashlib h = int hashlib.md5 name.encode .hexdigest , 16 levels = "Excellent", "Good", "Fair", "Poor" return {d: {"level": levels h i 3 % 4 , "reason": "offline mock score"} for i, d in enumerate DIMS } def gate score report : tot = sum LEVEL SCORE.get report.get d, {} .get "level", "Fair" , 2 for d in DIMS return tot / len DIMS 4 GATE THRESHOLD = 0.55 targets = s "skill url" for s in kw hits + vec hits if s "skill url" :3 \ or i "name" for i in inspected or "pdf-extractor", "chart-reader", "web-scraper" passed, scored = , for t in targets: rep = evaluate t via = "LLM" if rep is None: rep, via = mock eval str t , "mock" score = gate score rep scored.append t, score, via flags = " ".join f"{d :4 }={rep.get d,{} .get 'level','?' :4 }" for d in DIMS status = "PASS ✅" if score = GATE THRESHOLD else "FAIL ❌" print f" {via:4} {status} score={score:.2f} {textwrap.shorten str t ,46,placeholder='...' }" print f" {flags}" if score = GATE THRESHOLD: passed.append t print f"\n{len passed }/{len targets } skills passed the quality gate threshold={GATE THRESHOLD} ." banner "6 Analyze relationships and draw the Skill Graph" def analyze skills dir : if USE SDK and API KEY: try: return client.analyze skills dir=str skills dir except Exception as e: print f" analyze failed: {e r}" return None rels = analyze SKILLS DIR if not rels: names = i "name" for i in inspected or "PDF Parser", "Text Summarizer", "Chart Reader", "Web Scraper" while len names < 4: names.append f"Skill {len names }" rels = {"source": names 0 , "type": "compose with", "target": names 1 }, {"source": names 2 , "type": "similar to", "target": names 0 }, {"source": names 3 , "type": "depend on", "target": names 1 }, {"source": names 1 , "type": "belong to", "target": names 2 }, print " using offline mock relationships — set API KEY for real analysis " for r in rels: print f" {r 'source' } -- {r 'type' } -- {r 'target' }" try: import networkx as nx import matplotlib.pyplot as plt G = nx.DiGraph COLORS = {"similar to": " 4C9BE8", "belong to": " E8A14C", "compose with": " 6BBF59", "depend on": " D45D79"} for r in rels: G.add edge r "source" , r "target" , type=r "type" pos = nx.spring layout G, seed=42, k=1.2 plt.figure figsize= 9, 6 nx.draw networkx nodes G, pos, node size=2200, node color=" EDEDED", edgecolors=" 444" nx.draw networkx labels G, pos, font size=9 for et, col in COLORS.items : edges = u, v for u, v, d in G.edges data=True if d "type" == et if edges: nx.draw networkx edges G, pos, edgelist=edges, edge color=col, width=2, arrows=True, arrowsize=18, connectionstyle="arc3,rad=0.08" plt.legend handles= plt.Line2D 0 , 0 , color=c, lw=2, label=t for t, c in COLORS.items , loc="best", fontsize=8 plt.title "SkillNet — Skill Relationship Graph" plt.axis "off" ; plt.tight layout plt.savefig WORKDIR / "skill graph.png", dpi=130 plt.show print f" graph saved - {WORKDIR/'skill graph.png'}" except Exception as e: print f" graph drawing skipped: {e r}" We evaluate skills across five quality dimensions: safety, completeness, executability, maintainability, and cost awareness. We apply a quality gate to determine which skills meet a minimum score threshold, using mock scores when an API key is unavailable. We also analyze relationships between skills and visualize them as a Skill Graph using NetworkX and Matplotlib. banner "7 Skill-augmented agent planner" GOAL = "Analyze scRNA-seq data to find and validate cancer drug targets, then write a report" def llm decompose goal : if API KEY: try: payload = { "model": MODEL, "messages": {"role": "system", "content": "Decompose the user's goal into 3-6 short, ordered subtasks. " "Reply ONLY as a JSON array of strings, no prose, no markdown."}, {"role": "user", "content": goal}, , "temperature": 0.2, } r = requests.post f"{BASE URL}/chat/completions", headers={"Authorization": f"Bearer {API KEY}"}, json=payload, timeout=60 r.raise for status txt = r.json "choices" 0 "message" "content" txt = re.sub r"^ ?:json ?| $", "", txt.strip .strip subs = json.loads txt if isinstance subs, list and subs: return str x for x in subs except Exception as e: print f" LLM decompose failed - heuristic {e r} " return "acquire single-cell RNA-seq dataset", "preprocess and cluster cells", "identify candidate cancer target genes", "validate targets against pathway database", "generate a discovery report" def keywords for subtask : stop = {"the", "and", "a", "to", "of", "from", "into", "for", "with", "then", "an"} toks = w for w in re.findall r" a-zA-Z\- +", subtask.lower if w not in stop return " ".join toks :4 subtasks = llm decompose GOAL print f"GOAL: {GOAL}\n\nPLAN {len subtasks } steps :" plan = for i, st in enumerate subtasks, 1 : q = keywords for st hits = search q, mode="vector", limit=2, threshold=0.6 or \ search q, mode="keyword", limit=2 best = hits 0 if hits else None chosen = best "skill name" if best else " no skill found — fallback to base model " plan.append {"step": i, "subtask": st, "query": q, "skill": chosen} print f"\n Step {i}: {st}" print f" search '{q}' - {chosen}" + f" ⭐{best 'stars' }" if best else "" print "\nExecution order assembled pipeline :" print " " + " - ".join p "skill" .split 0 if p "skill" 0 = " " else "base-model" for p in plan banner "Tutorial complete" print textwrap.dedent f""" Recap: • Search keyword + vector ............ ran via {'SDK' if USE SDK else 'REST'} • Install GitHub - local ............ {len installed } skill s • Inspect SKILL.md metadata ............ {len inspected } parsed • Evaluate + quality gate .............. {len passed }/{len targets } passed {' LLM ' if API KEY else ' offline mock '} • Relationship graph ................... {len rels } edges - skill graph.png • Agent planner ........................ {len plan } steps mapped to skills Docs: https://github.com/zjunlp/SkillNet """ We build a skill-augmented agent planner around a complex scientific discovery goal. We decompose the goal into ordered subtasks, identify relevant skills for each step, and map those skills to an execution pipeline. We finish by printing a recap of the full workflow, including search, installation, inspection, evaluation, graph analysis, and planning. In conclusion, we created a complete SkillNet workflow that moves beyond simple skill search and demonstrates how skills can support structured agentic systems. We saw how SkillNet helps us discover useful capabilities, evaluate their quality, understand their relationships, and connect them to real task planning. It also remains practical because it runs even without an API key by falling back to offline mock evaluations, while still allowing deeper LLM-powered analysis when credentials are available. Also, we used SkillNet as a foundation for building modular, skill-driven AI agents that can plan, select tools, and organize execution more intelligently. Check out the Full Codes here. Also, feel free to follow us on and don’t forget to join our Twitter https://x.com/intent/follow?screen name=marktechpost and Subscribe to 150k+ ML SubReddit https://www.reddit.com/r/machinelearningnews/ . Wait are you on telegram? our Newsletter https://www.aidevsignals.com/ now you can join us on telegram as well. https://t.me/machinelearningresearchnews Need to partner with us for promoting your GitHub Repo OR Hugging Face Page OR Product Release OR Webinar etc.? Connect with us https://forms.gle/wbash1wF6efRj8G58 Sana Hassan, a consulting intern at Marktechpost and dual-degree student at IIT Madras, is passionate about applying technology and AI to address real-world challenges. With a keen interest in solving practical problems, he brings a fresh perspective to the intersection of AI and real-life solutions. - Sana Hassan - Sana Hassan - Sana Hassan - Sana Hassan