AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Humanizer

skill-dmitriyyukhanov-claude-plugins-humanizer · by DmitriyYukhanov

|

No reviews yet
0 installs
39 views
0.0% view→install

Install

$ agentstack add skill-dmitriyyukhanov-claude-plugins-humanizer

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-dmitriyyukhanov-claude-plugins-humanizer)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
2mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Humanizer? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Humanizer (EN + RU auto-routing)

Remove AI-generation patterns from text. This skill bundles two MIT-licensed upstream rulesets and routes the input to the right one based on language.

This file is a router. The actual humanization rules live in references/humanizer-en.md and references/humanizer-ru.md and are authoritative. Do not paraphrase, summarise, or "improve" them — load and follow.

Process

Step 1 — Receive input

Accept the text to humanize as: a pasted block, a file path (use Read), or the current selection. If the input is unclear, ask the user once.

Step 2 — Honour explicit override

Scan the user's request (case-insensitive) for an explicit language directive:

  • EN override — the request contains any of: as english, english only,

humanize in english, на английском, обработай как английский.

  • RU override — the request contains any of: as russian, russian only,

humanize in russian, на русском, обработай как русский.

If an override is present, skip step 3 and jump to step 4 with the chosen language.

Step 3 — Detect language

Count letters in the input (ignore digits, punctuation, whitespace):

  • cyr = number of letters matching [А-Яа-яЁё]
  • lat = number of letters matching [A-Za-z]
  • total = cyr + lat

Decide:

  • total = 0.6 → Russian.
  • cyr / total <= 0.1 → English.
  • Otherwise (mixed, Cyrillic ratio between 10 % and 60 % exclusive) → use AskUserQuestion with options

"Apply Russian rules", "Apply English rules", "Run both passes (RU then EN)".

Step 4 — Load the chosen ruleset

  • For English: Read references/humanizer-en.md
  • For Russian: Read references/humanizer-ru.md
  • For "both passes": load both. Execute RU first, then EN on the RU output.

The loaded file is authoritative. Treat its instructions as if they were yours.

Step 5 — Execute

Apply the diagnostic, rewrite, and audit process described in the loaded reference verbatim. Do not skip the audit pass. Do not invent new rules.

If the user provided a writing sample for voice calibration, pass it through to the chosen ruleset's voice-calibration section as-is, regardless of sample language.

Step 6 — Output

Use the output format specified by the loaded reference (typically: draft → audit → final → list of changes). Do not change the format.

Rules of the road

  • The router never edits the rulesets. Refresh from upstream is a separate

procedure documented in ../../../README.md.

  • The router never invents new patterns or merges English/Russian rules.
  • "Both passes" runs RU first because RU's hard bans include em-dash removal

(which already covers EN pattern #14) — running EN first would let dashes re-enter via RU's rewrites.

Attribution

This skill bundles MIT-licensed work by blader (English ruleset) and ilyautov (Russian ruleset). See ../../../NOTICE for full attribution and pinned upstream commits.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.