Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Explain Code · amariahak | DecimalAI

amariahak/Explain Code

SkillScore91Maintained

You are an expert coder that takes code and documentation as input and do your best to explain it.

codingby amariahak

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Coding pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/amariahak-explain-code/SKILL.md

decimalai skills pull amariahak-explain-code --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/amariahak-explain-code/SKILL.md https://app.decimal.ai/s/amariahak-explain-code/SKILL.md

Raw file: /s/amariahak-explain-code/SKILL.md ↗ · machine-readable summary at /s/amariahak-explain-code.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 55d ago

Yes — +77% lift-20% tokens0% turns91% of whole cases pass with skill (20/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-01	✗→✓	▲ Improved	-28%	0%
case-02	✗→✓	▲ Improved	-37%	0%
case-03	✗→✓	▲ Improved	-33%	0%
case-04	✗→✓	▲ Improved	-16%	0%
case-05	✗→✓	▲ Improved	-20%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/1/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

amariahak / atlarix-skills

skills/fabric-explain-code/SKILL.md

View on GitHub ↗

License Apache-2.0Imported 8/1/2026Version v1

Skill contents

SKILL.md

Explain Code

When to use this skill

You are an expert coder that takes code and documentation as input and do your best to explain it.

Source

Synced from https://github.com/danielmiessler/fabric/tree/main/data/patterns/explain_code/system.md.

IDENTITY and PURPOSE

You are an expert coder that takes code and documentation as input and do your best to explain it.

Take a deep breath and think step by step about how to best accomplish this goal using the following steps. You have a lot of freedom in how to carry out the task to achieve the best result.

OUTPUT SECTIONS

If the content is code, you explain what the code does in a section called EXPLANATION:.

If the content is security tool output, you explain the implications of the output in a section called SECURITY IMPLICATIONS:.

If the content is configuration text, you explain what the settings do in a section called CONFIGURATION EXPLANATION:.

If there was a question in the input, answer that question about the input specifically in a section called ANSWER:.

OUTPUT

Do not output warnings or notes—just the requested sections.

INPUT:

INPUT:

Illustrative Want to see what amariahak/Explain Code would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/4/2026, 08:59 UTC · 22 cases

Lift

+77%

91% vs 14%

Agent turns

0%

1.0 vs 1.0

Tokens

-20%

1,539.45 vs 1,931.73

fail → pass · 19pass → pass · 1pass → fail · 2

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 I need help understanding this Python function and a specific design choice in it. Could you explain what the code does, and also answer why it uses a set instead of a list for tracking seen items? def find_unique_users(log_entries): seen = set() for entry in log_entries: seen.add(entry['user_id']) return seen	fail→pass	10,139	6,464	-36%	1	1	0%	2,038	1,459	-28%	0	0	—
case-02 Here are the log outputs from a recent container vulnerability scan. Can you break down what these findings mean for our system's posture? Target: app-server:v1.2.0 CVE-2021-23017 \| Severity: HIGH \| Description: 1-byte memory overwrite in DNS resolver component	fail→pass	13,432	7,058	-47%	1	1	0%	2,415	1,531	-37%	0	0	—
case-03 Could you explain what these application configuration settings do so I understand how our connection limits are managed? max_connections = 100 idle_timeout = 300 connect_timeout = 10 statement_timeout = 5000	fail→pass	10,893	6,355	-42%	1	1	0%	1,935	1,303	-33%	0	0	—
case-04 Please analyze this Rust function that parses an string into a port integer: fn parse_port(val: &str) -> Option<u16> { val.parse::<u16>().ok() }	fail→pass	10,725	7,211	-33%	1	1	0%	2,081	1,749	-16%	0	0	—
case-05 Examine this Go concurrency construct and explain what it does: select { case msg := <-ch1: fmt.Println("received", msg) case <-time.After(2 * time.Second): fmt.Println("timeout") }	fail→pass	10,691	6,930	-35%	1	1	0%	1,549	1,234	-20%	0	0	—
case-06 Review this Nmap network scan output for host 10.0.0.15: PORT STATE SERVICE 22/tcp open ssh 80/tcp open http 3306/tcp open mysql Nmap done: 1 IP address (1 host up) scanned in 0.42 seconds	fail→pass	9,434	8,711	-8%	1	1	0%	1,606	1,466	-9%	0	0	—
case-07 Assess this Semgrep static analysis security scanner output for an Express app: [ERROR] js.express.security.audit.xss.direct-response-write Line 42: res.send(req.query.user_input) Severity: ERROR - Reflected Cross-Site Scripting (XSS)	fail→pass	10,385	16,438	+58%	1	1	0%	1,892	1,620	-14%	0	0	—
case-08 Break down what these Redis configuration directives control in memory management: maxmemory 4gb maxmemory-policy volatile-lru maxmemory-samples 5	fail→pass	9,950	6,368	-36%	1	1	0%	1,837	1,441	-22%	0	0	—
case-09 Explain the container resource allocation defined in this Kubernetes pod spec snippet: resources: requests: memory: "256Mi" cpu: "250m" limits: memory: "512Mi" cpu: "500m"	fail→pass	9,077	7,053	-22%	1	1	0%	1,682	1,641	-2%	0	0	—
case-10 Can you review this C code snippet and answer why ptr + 1 advances the memory address by 4 bytes instead of 1? int arr[5] = {10, 20, 30, 40, 50}; int ptr = arr; int val = (ptr + 1);	fail→pass	8,361	4,898	-41%	1	1	0%	1,762	1,276	-28%	0	0	—
case-11 Here is an AWS GuardDuty security alert log. What does this alert indicate, and what is the severity rating? FindingType: UnauthorizedAccess:EC2/SSHBruteForce Severity: 5.0 Resource: i-0a123456789abcdef Description: 1,420 failed SSH login attempts from IP 198.51.100.44	fail→pass	8,621	6,139	-29%	1	1	0%	1,598	1,427	-11%	0	0	—
case-12 What does this HAProxy configuration section do, and how does it handle backend server health checks? backend web_servers balance roundrobin option httpchk GET /health server s1 192.168.1.10:8080 check inter 2000 fall 3 rise 2 server s2 192.168.1.11:8080 check inter 2000 fall 3 rise 2	fail→pass	10,258	9,846	-4%	1	1	0%	2,043	1,883	-8%	0	0	—
case-13 Detailed analysis requested for this Python decorator pattern: from functools import wraps def retry(retries=3): def decorator(func): @wraps(func) def wrapper(args, kwargs): for _ in range(retries): try: return func(args, **kwargs) except Exception: pass raise RuntimeError("Failed after retries") return wrapper return decorator	fail→pass	18,626	8,450	-55%	1	1	0%	2,914	1,799	-38%	0	0	—
case-14 Analyze this Trivy vulnerability scanner result for an Alpine Linux image: Library: libcrypto3 (3.0.8-r0) Vulnerability ID: CVE-2023-0464 Severity: HIGH Installed Version: 3.0.8-r0 Fixed Version: 3.0.8-r1 Title: OpenSSL policy validation resource exhaustion	fail→pass	10,974	5,369	-51%	1	1	0%	2,028	1,220	-40%	0	0	—
case-15 Clarify the network behavior configured by these Linux sysctl kernel settings: net.ipv4.tcp_syncookies = 1 net.ipv4.tcp_max_syn_backlog = 2048 net.ipv4.tcp_synack_retries = 2	fail→pass	12,949	6,936	-46%	1	1	0%	2,099	1,492	-29%	0	0	—
case-16 Walk through this Java class and answer why incrementing count causes race conditions without synchronization: public class Counter { private int count = 0; public void increment() { count++; } }	fail→pass	16,226	5,910	-64%	1	1	0%	1,896	1,324	-30%	0	0	—
case-17 Evaluate this Suricata IDS alert output generated from HTTP web traffic: [1:2001219:15] ET WEB_SERVER SQL Injection Attempt -- SELECT FROM Classification: Web Application Attack Priority: 1 IP: 203.0.113.88 -> 192.168.1.50	fail→pass	13,769	7,206	-48%	1	1	0%	2,511	1,648	-34%	0	0	—
case-18 Determine the TLS security posture defined by this Nginx configuration snippet: ssl_protocols TLSv1.2 TLSv1.3; ssl_ciphers HIGH:!aNULL:!MD5; ssl_prefer_server_ciphers on; ssl_session_cache shared:SSL:10m;	fail→pass	13,012	10,217	-21%	1	1	0%	2,525	2,138	-15%	0	0	—
case-19 Describe how Promise.allSettled operates in JavaScript and answer how its behavior differs from Promise.all when a fetch fails: const results = await Promise.allSettled([ fetch('/api/v1/users'), fetch('/api/v1/posts') ]);	fail→pass	10,496	6,871	-35%	1	1	0%	1,989	1,552	-22%	0	0	—
case-20 Write a refactored version of this Python function to use list comprehensions for better readability. def get_evens(numbers): evens = [] for n in numbers: if n % 2 == 0: evens.append(n) return evens	pass→fail	3,621	4,269	+18%	1	1	0%	457	975	+113%	0	0	—
case-21 Generate a Bash shell script to update Nginx to patch CVE-2021-23017 on an Ubuntu 20.04 server.	pass→pass	13,454	9,040	-33%	1	1	0%	2,701	1,911	-29%	0	0	—
case-22 Create a docker-compose.yml file configured for a PostgreSQL 15 database with a max_connections setting of 200.	pass→fail	4,761	9,523	+100%	1	1	0%	940	1,779	+89%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +77 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 13.6%
Pass rate with the skill: 90.9%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 2
Skill version measured: v1
Run completed: 8/4/2026, 08:59 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-01

Prompt

I need help understanding this Python function and a specific design choice in it. Could you explain what the code does, and also answer why it uses a set instead of a list for tracking seen items?

def find_unique_users(log_entries):
    seen = set()
    for entry in log_entries:
        seen.add(entry['user_id'])
    return seen

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run91

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns-29% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 54 days ago

Other measured skills in the registry, with their headline benchmark lift.