muvon/octomind-tap/social-devto

Claude Code and agent skill by muvon/octomind-tap

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Write and revise DEV articles, discussions, comments, and launch or proof material in the author's documented voice. Activate for dev.to drafts, publishing packages, series, challenge entries, or company cross-posts; apply current disclosure and promotion rules before preparing copy for publication.

Install

npx vskill@latest install muvon/octomind-tap/social-devto

SKILL.md

Overview

Write something a developer can use or answer, in the voice of the person accountable for it. Apply supplied facts and proof first, the author's documented voice next, and platform register as the fallback; respect publication policy throughout. Use content-voice for generic editing and the rules here for DEV structure and publishing decisions.

Check the requested surface first: DEV's current AI guidelines allow qualifying assisted articles but prohibit AI-generated comments, with exceptions for basic translation, grammar/syntax improvement, and assistive technology. For a new comment, provide factual notes for independent human authorship; do not output a polished reply to paste. For an allowed language edit, preserve the human's argument and don't add generated experience or advice (official, DEV AI Guidelines, checked 2026-09-05).

Mechanics and evidence

  • Treat discovery as personalized: follows and reactions interact with semantic similarity, quality, and time decay. Don't promise reach from a tag recipe (official, DEV Feed, 2026-05).
  • Use the current reaction vocabulary: Love, Unicorn, Wow, Well Done, and Hot Take (official, DEV Engagement, 2026-09). Don't assign bookmarks a strongest-signal weight or treat reactions as purchase intent.
  • Recognize Community Gems as curator endorsements that affect author trust and discovery, including a Curated feed; weighting is evolving (official, DEV Gems, 2026-09). Earn recognition through useful work; don't solicit reciprocal awards (directional).
  • Treat public feed-code inspections as models, not production measurements. Comments, follows, language, and age affected the inspected variant; no universal comment multiplier or visibility window follows from it (measured, Loibner, n=single checked-in variant, 2026-07).

Open reference/publishing-evidence.md when checking editor fields, planning an AMA or challenge, choosing a timing experiment, or citing performance. It contains the evidence limits and source register; don't turn its account case studies into targets.

Format and publishing decisions

Choose by the reader's job. These shapes and qualitative length bands are editorial guidance, not platform post-type guarantees (directional).

JobShape and length bandRequired substance
Reproduce a techniqueArticle; full walkthrough to reference-lengthWorking example, setup conditions, explanation, failure boundary
Compare real experiencediscuss; brief context to short argumentState the unresolved decision and your partial answer; ask an answerable question
Evaluate a built artifactshowdev; compact demonstration to full walkthroughShow the artifact and how to try it; disclose ownership and limitations
Follow a continuing investigationSeries; self-contained article per installmentRecurring problem, new evidence in each entry, useful stopping point
Ask about a person's expertiseAMA; brief invitationVerifiable remit, excluded topics, actual reply availability
Enter a challengeAnnouncement's required template; sufficient detail for judgingPrompt fit, eligible build, reproducible submission evidence

Series navigation appears after the second entry (official, DEV Writing, 2026-09). Don't split an explanation merely to multiply posts. A listicle earns its format when entries help readers choose or express the author's specific humor; remove interchangeable filler (directional).

Front-load the technical situation or defensible claim in the title. A genuine question can be the title or opening when its scope is already clear; add context where readers need it. Replace a rhetorical question only when it delays or exaggerates the actual point (directional).

Tags and editor package

Use at most four relevant tags (official, DEV Editor, 2026-09). Treat the following allocation as a selection method, not a distribution formula (directional).

CandidateKeep whenDrop or replace when
Broad topic, such as programming or webdevIts current tag description fits the articleAnother tag describes the intended reader more precisely
Format, such as discuss or showdevThe draft actually invites discussion or demonstrates a projectThe article is a tutorial or opinion without that job
Niche, such as rust, postgres, or accessibilityThe implementation materially uses that topicIt appears only in passing
Language-community tagCurrent guidance and recent posts fit the languageYou are guessing its purpose or copying another language's tag recipe

Check each candidate's live page, description, and recent posts. Record visible follower counts with capture dates; mark unavailable counts unknown. Among equally relevant tags, compare current audience size; don't maintain a supposedly permanent biggest-tag list (directional).

Prepare an editor checklist for title, published, tags, cover_image, canonical_url, series, and description; inspect the current editor before mapping the package to front matter. Keep publication disabled during review, quote YAML strings when needed, and preview the rendered article. The reference separates verified features from unverified field constraints (directional).

DEV supports Markdown and documented Liquid embeds including GitHub, CodePen, and Twitter (official, DEV Editor, 2026-09). Add language hints to article code fences. Preview every embed; provide a useful text link if rendering fails. Choose a real screenshot when it explains the work, with descriptive alt text and a caption identifying its context. Inspect the actual cover crop; don't promise a click lift (directional).

Cross-post complete useful work with canonical_url pointing to the original; DEV supports canonical attribution and RSS imports (official, DEV Writing, 2026-09). Repair relative links and missing assets. Don't describe canonical attribution as guaranteed SEO protection (directional).

Launch and proof posts

For campaign planning, record audience, buying situation, promise, sourced proof, desired action/destination, stage, and disclosures. Use the author's notes and voice samples when available; missing samples alone don't block permitted work. Choosing no CTA is valid when no reader action is needed. Narrow unsupported claims or request essential missing evidence instead of manufacturing experience; the authorship gate below still applies (directional).

Apply the policy gate before drafting promotional copy: the currently linked guidelines prohibit AI-assisted or generated articles that promote a business, program, or course, or primarily build a personal brand or social presence. They also prohibit using generated educational information the author does not understand. AI-written launch copy from this workflow isn't cleared for DEV publication. Provide an evidence brief for independent human authorship, or qualifying nonpromotional education within the author's knowledge; removing a CTA or selecting a label doesn't erase promotional purpose. The August disclosure announcement still links these rules and says revision is forthcoming; it doesn't repeal them (official, DEV AI Guidelines and DEV AI Disclosure, checked 2026-09-05).

For policy-eligible work, choose the following anatomy and qualitative length band (directional):

ShapeAnatomyLength band
Launch dayDeveloper problem → artifact in use → design choice → limitation → relevant destinationCompact announcement to walkthrough
Demo or artifactInput → observable output → reproduction instructions → unsupported caseShort demonstration to full tutorial
Customer proofBefore/after under comparable conditions → source and permission → confounders → who can reuse itFocused case study
Founder decisionActual disputed choice → evidence considered → cost accepted → next unresolved testShort update to technical essay
Objection answerReader's objection → direct answer → reproducible evidence → remaining limitBrief answer to worked comparison
RecapOriginal promise → observed outcome → unresolved issue → useful next artifactCompact review to reference article

For numerical proof, name the metric and unit, baseline and comparison values, sample size, measurement window, workload or cohort, exclusions, and source with reuse permission. For rates, retain the numerator and denominator; distinguish relative change from percentage-point change. Keep these bounds beside the result. If essential inputs are missing, omit the numerical claim or leave it explicitly unpublishable (directional).

Use the strip-test: removing the product name and CTA should leave something worth learning. DEV recommends complete educational articles and warns against over-promotion (official, DEV Organization Guide, 2026-09). Bring niche expertise through a familiar developer problem; open the reference for the bounded case evidence behind that choice.

Choose an individual byline for personal work. For company work, credit the real contributor within the Organization, complete their profile, and use canonical attribution for company-blog reposts; DEV recommends human profiles and supports Organization sidebar CTAs (official, DEV Organization Guide, 2026-09). State actual role and subject expertise in the bio; avoid a generic founder pitch (directional).

Put the necessary demo or source link beside the evidence and the conversion CTA in the Organization sidebar or a brief outro. Avoid repeating the same self-link; allow additional owned links only when each supports a distinct claim. Don't hide the destination in a comment to chase an assumed reach benefit. State affiliation beside product claims and sponsorship near the opening; confirm any required placement label in the publishing workflow (directional).

For destination attribution, use consistent utm_source, utm_medium, and utm_campaign; distinguish article and sidebar links with utm_content (official, Google URL Builder, 2026-09). Keep attribution separate from claims that DEV caused a sale (directional).

DEV recommends relevant commenting and following, plus Welcome and discussion participation (official, DEV Organization Guide, 2026-09). Participate where you can contribute without pitching. Have collaborators disclose their relationship and add independent substance; exclude vote rings, scripted applause, disguised employees, and unsolicited link drops. These are community-integrity guardrails, not a claimed DEV penalty formula (directional).

Use this illustrative launch-week plan only when the policy gate passes. T offsets are planning labels, not measured optimal days (directional).

Illustrative offsetWork and reply planObserve
T-7 to T-1 (illustrative plan; directional)Join relevant discussions; prepare proof and preview; skip empty teasersRecurring developer questions and useful profile context
T0 (illustrative plan; directional)Publish the native demo; author covers the first hours for setup failures and substantive questionsViews, qualified technical comments, missing instructions
T+1 to T+3 (illustrative plan; directional)Correct the post; answer the strongest objection if new evidence warrants itReproduction reports, product-fit questions, source-tagged visits
T+4 to T+7 (illustrative plan; directional)Publish a technical follow-up or bounded recap; continue replies on the originalReturning commenters, follows, Gems, downstream activation

Voice on this platform

Use proofread, approachable technical prose as the fallback register. Preserve the author's normal contractions and terminology. First-person claims require their records or testimony. Keep reproducible code exact; never inject typos, fake edits, lowercase camouflage, or dropped articles (directional).

This is the start of the file. Read the full SKILL.md on GitHub.

Security

  • Security scanPassed, 95 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
5

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/muvon/octomind-tap/social-devto/badge)](https://verified-skill.com/skills/muvon/octomind-tap/social-devto)

Something wrong with this skill or its page? Tell us.