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 synaptiai/synapti-marketplace --skill prose-claritygit clone --depth 1 https://github.com/synaptiai/synapti-marketplaceWrote 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/synaptiai/synapti-marketplace/prose-clarity)<a href="https://agentmods.dev/skills/synaptiai/synapti-marketplace/prose-clarity"><img src="https://agentmods.dev/badge/skills/synaptiai/synapti-marketplace/prose-clarity/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/synaptiai/synapti-marketplace/prose-clarity"><img src="https://agentmods.dev/badge/skills/synaptiai/synapti-marketplace/prose-clarity.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 | $0.00146 | $0.01366 |
| Opus 5 | $0.00073 | $0.00683 |
| Sonnet 5 | $0.00029 | $0.00273 |
| Haiku 4.5 | $0.00015 | $0.00137 |
Grade A, and why
prose-clarity 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 8d 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 — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Prose Clarity
Owns the sentence- and paragraph-level form of drafted prose — never its truth, never its structure, never its claim state.
Iron Law
THE LINTER CATCHES FORM, NEVER TRUTH — AND A REQUIRED HEDGE ON AN I-STATE CLAIM IS NEVER A VIOLATION, IN ANY MODE.
A sentence that must carry [EV-####], keep Inferred:/Unknown:/Recommendation:, or preserve a limitation that keeps it true is correct at whatever length that takes. This skill fixes the form of slop. It cannot make a hollow paragraph true, and it never overrides an Iron Rule from another skill.
Classification is by passage, not by document
Any line inside a numbered or bulleted step sequence, a runbook or procedure block, or a command caption resolves to strict mode: 20-word sentence cap, full rule set. Everything else — narrative rationale, executive summary, due-diligence prose — resolves to STE-flavored: 25-word cap, relaxed dictionary, same active-voice and no-phrasal-verb discipline. This is mechanically detectable from list markers and fenced/table syntax, and it handles a single document that mixes both, which most of them do.
Hard categories — apply in every mode, always block
Banned marketing adjectives · banned phrasal verbs · semicolons · banned filler and hedge phrases (minus the carve-out below) · sentence-length cap for the resolved mode · paragraph cap (more than six sentences). Full lists: references/prose-style-and-vocabulary.md.
Advisory categories — reported, never block
Passive voice, nominalization, em-dash count. Both grammar checks are heuristic and false-positive-prone — this plugin's own reference docs use em-dashes constitutively, so em-dash must never gate release. Advisory hits feed dossier-pass-c-audience's Step 5 and the scorecard's Dimension 10 as context for whether a hard-category fix was genuine or superficial, not as findings on their own.
The epistemic-hedging carve-out
references/source-authority-and-claim-states.md requires specific hedge phrases as structural markers for Interpreted-state claims: "This suggests…", "The most likely reading is…", "Inferred from…", and the prose prefixes Inferred:, Unknown:, Recommendation:. These are the package's epistemic-honesty mechanism, not slop. A line opening with one of these markers is exempt from the hedge-phrase and sentence-length rules, in every mode, unconditionally. The full allow-list lives in references/prose-style-and-vocabulary.md — check it before ever touching a flagged hedge.
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.
- 8d ago First seen · 71 lines · 146 tokens per session scan A 1de733bf1149
prose-clarity is a skill published in the GitHub repository synaptiai/synapti-marketplace (6 stars, last pushed yesterday), licensed Apache-2.0. It adds 146 tokens to every session and 1,366 once invoked, about $0.0007 per session on Opus 5. 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 2026-09-03.
Other skills, from other repositories
pdf-cascade
Acquire PDFs for bibliographic references via a strict 8-source cascade (Crossref OA → arXiv → OpenAlex → Unpaywall → HAL → CORE → archive.org → WebSearch queue; optionally Sci-Hub + Anna's Archive, plus Anna's via a windowed browser, in opt-in mode). Each acquired PDF is validated against expected author/title/year…
document-writer
Use when writing or improving technical documentation, API references, runbooks, onboarding guides, READMEs, changelogs, ADRs, release notes, or any content that ships to engineers or end users. Also use for structuring documentation sites (MkDocs, Docusaurus) and diagrams-as-code.
kb-lint
Health-check the knowledge bank for orphan documents, stale content, broken WikiLinks, index drift, and missing frontmatter. Generates a severity-graded report. Use this skill whenever the user mentions lint, health check, KB maintenance, stale docs, broken links, orphan pages, missing references, knowledge bank…
create-or-update-concepts
Scan and analyze the project codebase to create or update concept files in agents-context/concepts/ (organized by domains/, source/, shared/) and update the README index and Load-When Cheatsheet.
step2-scope-tasks
Break a specification into ordered task groups with explicit context-awareness directives.
step1-write-spec
Gather requirements through structured Q&A, then formalize into a specification document.