Skills/npm/humanizer-lint

humanizer-lint@1.6.0

CAUTION· install only

Publisher claims: "A local, zero-dependency prose linter that flags stiff, AI-sounding writing (stock phrases, flat sentence rhythm, repeated wording) and tells you what to revise. A style checker, not an AI detector." Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry npm (MCP server)Publisher yanstd888 · publishing for 15 daysInstalls 2,694Last detonated 6d ago
What would fix thisWatchDispute verdict

Coverage

Nothing executable in the bundle ran; the trace covers the install step only.
Tested
every file read, file written and host contacted by what ran
Not tested
use of the skill by an agent

Is [email protected] safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (1 scenario; nothing executable in the bundle ran; the trace covers the install step only) on 2026-11-08 and scored it 63/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Prompt injection; Credential pass-through configuration; Prompt injection. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 0 s · 1 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
0.0sprocessscript backend runs setup only
0.0sprocessscript backend runs setup only
0.0sprocessno install scripts to re-run

Static findings

CRITICAL
Prompt injectionSKILL.md:33 · Prompt injection — known jailbreak technique reference… menandai tulisan kaku ala AI (frasa klise, ritme kalimat datar, kata berulang) dan menunjukkan bagian yang perlu direvisi. Pemeriksa gaya, bukan detektor AI. #…Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionrules/id-slop.json:33 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file{"id":"ID006","type":"SLOP_PHRASE","pattern":"\\bperlu dicatat bahwa\\b","fixInstruction":"Hapus pengantar dan mulai dari informasi intinya."},Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionREADME.md:43 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file…sentence rhythm. ID · humanizer bahasa Indonesia: perbaiki tulisan AI yang kaku dan audit ritme kalimat. ES · humanizador de texto IA y corrector de estilo y ritmo…Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
HIGH
Credential pass-through configurationscripts/preserve-check.mjs:60 · Credential pass-through configuration — agent relays sensitive credentials// ponytail: two-pass token extraction with delimiter stripping; digit groupingFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST02
CRITICAL
Prompt injectionSKILL.md:10 · Prompt injection — known jailbreak technique referencelebih natural, ubah teks jadi lebih natural, cek ritme kalimat dan kataFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionbenchmark/eval-baseline.json:1034 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file…atau masih dianggap misteri, tapi hal tersebut dapat meningkatkan fleksibilitas dan produktivitas sebagai seorang teknisi secara jelas. Halaman ini merupakan kumpu…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST01
LOW
Prompt injectionrules/id-slop.json:17 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file…":"ID052","type":"FORCED_TRIAD","pattern":"\\b[\\p{L}]{3,},\\s+[\\p{L}]{3,},\\s+dan\\s+[\\p{L}]{3,}\\b","fixInstruction":"Review this pattern in context; revise on…Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionbenchmark/eval-baseline.json:880 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file…in this library are the property of their respective owners. - Algoritma Nazief dan Adriani - Asian J. 2007. ___Effective Techniques for Indonesian Text Retrieval_…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST01
HIGH
Hidden zero-width characters detectedbenchmark/eval-baseline.json:1160 · Hidden zero-width characters detected — may conceal content…", "text": "Candi Prambanan atau Candi Roro Jonggrang (bahasa Jawa: ꦕꦤ꧀ꦝꦶ​ꦥꦿꦩ꧀ꦧꦤꦤ꧀, translit. Candhi Prambanan) adalah kompleks candi Hindu terbesar di In…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
MEDIUM
File type combined with upload actionscripts/eval.mjs:26 · File type combined with upload action — agent may exfiltrate local data filesif (values.json && values.table) throw new Error('Choose one output format');Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
LOW
Prompt injectionrules/id-slop.json:32 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file{"id":"ID005","type":"SLOP_PHRASE","pattern":"\\bpenting untuk diingat\\b","fixInstruction":"Hapus penanda kepentingan dan tulis fakta yang perlu diingat."},Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionrules/id-slop.json:11 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file"stopwords": ["yang", "dan", "di", "ini", "untuk", "dengan", "dari", "pada", "dalam", "tidak", "akan", "atau", "juga", "oleh", "adalah", "sebagai", "bahwa", "dapat", "karena", "tersebut"],Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionREADME.md:82 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file- **ID:** Untuk ubah tulisan AI jadi natural dan hilangkan kata klise AI, gunakan editor gaya bahasa dan ritme kalimat untuk audit artikel SEO dan dokumentasi teknis tanpa mengubah fakta.Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionbenchmark/eval-baseline.json:820 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file…cal-doc", "split": "dev", "wordCount": 404, "provenance": "Dan Abramov and Redux Contributors, Redux Introduction: Core Concepts; commit 5c8b5…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST01
LOW
Prompt injectionrules/id-slop.json:29 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file{"id":"ID002","type":"SLOP_PHRASE","pattern":"\\bdi tengah pesatnya\\b","fixInstruction":"Hapus pembuka umum dan sebutkan perubahan yang dimaksud."},Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
HIGH
Hidden zero-width characters detectedbenchmark/eval-baseline.json:1090 · Hidden zero-width characters detected — may conceal content… maka penjajahan di atas dunia harus dihapuskan, karena tidak sesuai dengan peri­kemanusiaan dan peri­keadilan.\n\nDan perjuangan pergerakan kemerdekaan Indonesi…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
MEDIUM
File type combined with upload actionbenchmark/eval-baseline.json:880 · File type combined with upload action — agent may exfiltrate local data files…alled with [Composer](https://getcomposer.org). Add sastrawi into your `composer.json`: ```json { \"require\": { \"sastrawi/sastrawi\": \"*\" } } ``` Then run `composer install` or `composer update` from `command line`. Usage ------ Copy t…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
MEDIUM
npx invocation of non-well-known packageSKILL.md:45 · npx invocation of non-well-known package — may execute untrusted codeDo not call network tools or upload the user's text. `npx humanizer-lint` is for manual CLI or CI use outside this agent workflow.Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
CRITICAL
HTML comment carrying imperative instructions to the agentREADME.md:282 · HTML comment carrying imperative instructions to the agenthumanizer-ignoreFix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionREADME.md:41 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file… menandai tulisan kaku ala AI (frasa klise, ritme kalimat datar, kata berulang) dan menunjukkan bagian yang perlu direvisi. Pemeriksa gaya, bukan detektor AI. *…Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
LOW
Prompt injectionrules/ms-slop.json:8 · Prompt injection — known jailbreak technique reference · informational: not an instruction file · informational: not an instruction file"dan",Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01

Cost to the agent

Instructions loaded
≈2,189 tok
1 file · 44 in bundle

Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.

Known vulnerabilities

No advisories in OSV for humanizer-lint@1.6.0 as of 6d ago. This is the package itself, not its dependencies.

What would make this Verified

17 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%
  1. Text that tells the agent to ignore its instructions≈ +9 on the totalDelete phrases like "ignore previous instructions", "override", "bypass" from SKILL.md and tool descriptions. Describe what the skill does; do not instruct the agent about other rules.
  2. Unclaimed publisher≈ +8 on the totalClaim yanstd888: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
  3. Hidden instructions≈ +6 on the totalRemove HTML comments, zero-width characters and white-on-white text that carry instructions. Everything the agent is meant to read should be visible to the person reading the file.
  4. Credential pass-through configuration≈ +4 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
  5. Hidden zero-width characters detected≈ +4 on the totalPut hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
  6. New publisher≈ +3 on the totalTime clears this on its own after 14 and 30 days. Claiming the handle (+50) is the fastest way past it.
Publish a new version to npm, then queue it here. Not your skill? Claim the handle to get verdict changes by email.Claim yanstd888Propose a patch to the maintainer