Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add petar-djukic/writing-skills --skill voice-criticgit clone --depth 1 https://github.com/petar-djukic/writing-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/petar-djukic/writing-skills/voice-critic)<a href="https://agentmods.dev/skills/petar-djukic/writing-skills/voice-critic"><img src="https://agentmods.dev/badge/skills/petar-djukic/writing-skills/voice-critic/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/petar-djukic/writing-skills/voice-critic"><img src="https://agentmods.dev/badge/skills/petar-djukic/writing-skills/voice-critic.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.01460 |
| Opus 5 | $0.00000 | $0.00730 |
| Sonnet 5 | $0.00000 | $0.00292 |
| Haiku 4.5 | $0.00000 | $0.00146 |
Grade A, and why
voice-critic scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 118 lines — stays where its author put it; the contents beside it link to each section on GitHub.
name: voice-critic description: >- Cold, read-only gatekeeper against writing-voice/voice-constitution.md: five per-dimension verdicts with flagged spans — stance, theory-of-mind device, disproportion (locked-span check), computed marker profile against idiolect targets, and a snark audit on the constitution's L0-L5 scale with receipt-first, safe-enemy, and per-form density-cap hard rules. Never edits; flags route to the author gate. Triggers: voice critic, does this sound like me, snark audit, constitution check, voice gate, critique voice, off-voice check.
Voice Critic (read-only gatekeeper)
filter-tells detects "sounds like a machine". This critic detects two
different failures: "doesn't sound like Petar" and "went too far". The
rubric is writing-voice/voice-constitution.md — discovered by the
standard walk-up, refused when absent, because the critic's taste is that
file, not anything built in. Rates come from idiolect.yaml, the single
source of truth the constitution itself defers to.
Read-only is the contract. scripts/voice_critic.py takes a document
and produces a report; it never modifies the input, and the tests assert
input bytes unchanged. That is why it may run after the terminal
inject-vernacular stage — its home is the caller's review phase (GH-208),
after critic-panel — and equally at any earlier gate; nothing in it
assumes pipeline position. It is not a stage of the humanize chain.
The critic flags; the author adjudicates. It replaces self-monitoring with a checklist, not with enforcement: over-ceiling snark is flagged, never auto-deleted, and every flag carries the quoted span and line range so the gate read is a yes/no per item.
The five dimensions
| dimension | how | what it asks |
|---|---|---|
| stance | judged | Expert-watching, curiosity not contempt (§1)? At the bench, not the balcony? |
| tom-device | screened + judged | Model stated → predictions derived → tested against the artifact (§2)? The deterministic screen finds candidate evidence for each of the three parts; the judge (or the author) says whether the device is executed. |
| disproportion | computed | One declared overrun present and protected? Span locks are the declaration mechanism, so this reads the drivers' lock report. |
| marker-profile | computed, never judged | idiolect.yaml regex rates against essay targets, ±30% tolerance — the same arithmetic inject-vernacular applies as operators, run here as a check. |
| snark-audit | judged instances, computed rules | Every instance scored L0–L5; then the three hard rules run mechanically over the instances. |
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago Changed · +2 lines 15d833b4a01a
- 12d ago First seen · 116 lines · 0 tokens per session scan A 05ab712baaf4
voice-critic is a skill published in the GitHub repository petar-djukic/writing-skills (4 stars, last pushed 8d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,460 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
humanizer
Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague…
structural-humanizer
Remove the discourse-level (structural) signs of AI writing that survive surface editing: stated lessons and moral-of-the-story closers, tidy single-track arcs, embodied-emotion performance ("chest tightened"), vague allusions instead of named references, unbroken linear structure, and shape convergence across pieces.…
story-deslop
A writing editor for Chinese web novels that detects writing patterns often associated with AI-generated text and makes the prose feel more natural.
story
A Chinese-language routing tool for fiction writers that directs requests to tools for outlining, writing, analysis, reviews, covers, importing novels, browser control, and project management.
review-work
Post-implementation gate review: run manual QA on the real surface yourself, then launch ONE gate reviewer (never a panel) to audit goal, constraints, code quality, security, missed context, and QA evidence. Use before a PR handoff or when the user explicitly asks to review completed work.
critical-code-reviewer
Rigorously review code or pull requests for correctness, security, accessibility, maintainability, tests, and edge cases. Use when users request a critical code review, want a guided walkthrough of findings, need implementer-facing feedback, or want to prepare, create, or submit a GitHub pull request review.