build-in-public
Best practices for sharing engineering work publicly — what to post, what to withhold, and channel-specific guidance
Skill by alinaqi
Chalk Labs
Need help deploying build-in-public?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/alinaqi-build-in-public
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/build-in-public
What's inside
- Skill name
build-in-public- Description
- Best practices for sharing engineering work publicly — what to post, what to withhold, and channel-specific guidance
- License
- MIT
- Instructions
- 139 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Build in Public — Best Practices
## Philosophy
Build in public isn't marketing. It's letting people watch you work. The best posts feel like you're narrating your thought process to a friend who's also a senior engineer. No hype sludge. No "I'm excited to announce." Just: here's what I built, here's why it matters, here's what I learned.
## What to Share (and What Not To)
**Share:**
- Technical decisions and the reasoning behind them ("Chose SQLite over Postgres because...")
- Architecture insights ("Here's how the 9-tier routing pipeline works")
- Failures and what you learned ("Spent 3 hours debugging a race condition. Root cause: ...")
- Before/after metrics ("Compaction death spirals: 26 → 0 after Mnemos")
- Open source releases with context, not just links
- Counter-intuitive findings that surprised you
**Never share:**
- Revenue numbers, user counts, valuation
- Customer names or identifiable client details
- Internal URLs, API keys, credentials
- Anything covered by NDA
- Roadmap promises you might not keep
- "We're hiring" disguised as content
## Channel-Specific Best Practices
### LinkedIn
**Audience:** Engineers, CTOs, founders who build. They scroll between meetings looking for something that makes them think.
**Format:**
- 1-3 paragraphs, 800-2000 characters sweet spot
- Lead with the insight, not the context
- Use line breaks generously — wall of text kills engagement
- One clear takeaway per post
- No external links in first paragraph (LinkedIn penalizes off-platform clicks)
**Tone:** Confident, teaches something. You're the senior engineer explaining your approach to a peer. No jargon without explanation. If you used a technique others might not know, explain it briefly.
**When to post:** Tuesday-Thursday, 8-10 AM in target timezone. Avoid weekends and Monday mornings (everyone's catching up).
**What works:**
- Technical deep dives with concrete code examples
- "How I built X" narratives with architecture diagrams
- Lessons learned from failures (these outperform success stories 3:1)
- Opinionated takes on industry trends (but only if you have data)
**What flops:**
- "Excited to announce" press releases
- Pure product updates without technical insight
- Motivational content without substance
- Posts longer than 2500 chars without strong hook
### X (Twitter)
**Audience:** Developers who ship. They scroll fast and judge faster. You have one sentence to earn their attention.
**Format:**
- 280 characters max
- No threads unless the insight genuinely needs 3+ posts
- Lead with the counter-intuitive or surprising element
- One idea per post. If you have two ideas, make two posts.
- Screenshots need alt-text describing what's shown
**Tone:** Sharp, opinionated, zero filler. Imagine you're texting a builder friend. If it sounds like marketing, delete it.
**When to post:** Tuesday-Friday, 9-11 AM or 2-4 PM in target timezone. Weekends can work for developer audience (they're building side projects).
**What works:**
- One-sentence technical insights ("The difference between a good API and a great one is error messages.")
- Before/after comparisons with metrics
- "Just shipped X. Here's the one thing that surprised me."
- Asking genuine technical questions (engagement bait backfires)
**What flops:**
- Hashtag stuffing
- Threads that could be one post
- Generic "hot take" without personal experience
- Posting links without context
## Content Calendar Rhythm
**Daily (if you have something to say):**
- One X post about what you're working on or learned today
**Weekly:**
- One LinkedIn post: deeper technical insight or project milestone
**Per event (triggered by plugin):**
- PR merged → LinkedIn within 24h, X same day
- Feature shipped → both channels, LinkedIn first, X 90 min later
- Review passed → X only (architecture insights are punchy)
- Major release → LinkedIn deep dive + X announcement
## Anti-Patterns to Avoid
1. **The "we" trap** — Solo builders using "we" sounds insecure. Use "I" unless you're actually a team.
2. **Engagement bait** — "Agree?" or "Thoughts?" at the end of every post reads as desperate. Let the insight stand alone.
3. **Posting without building** — If you haven't shipped in 2 weeks, don't post. Your content should be a byproduct of your work, not a substitute for it.
4. **The LinkedIn bro voice** — "I'm humbled and honored to share..." Delete immediately. You're not accepting an award.
5. **Over-polishing** — A post that sounds like it went through 5 rounds of editing reads as corporate. Ship the draft.
6. **Ignoring replies** — If someone takes time to engage, reply within 24h. The conversation in comments often outperforms the original post.
## Measuring What Works
Track these signals (Buffer, LinkedIn analytics, X analytics):
- **Impressions** — how many people saw it
- **Engagement rate** — (likes + comments + reposts) / impressions
- **Profile visits** — did the post drive people to learn more?
- **Inbound** — DMs, connection requests, or emails referencing specific posts
A good LinkedIn post: 3-5% engagement rate. A great one: 8%+. On X, anything above 2% is solid for technical content.
## Plugin Integration
The build-in-public plugin follows these practices automatically:
```yaml
# What gets shared vs skipped:From alinaqi/maggy, MIT.
More from alinaqi/maggy
- 0.761score
model-routing
alinaqi
9-tier model routing system with cascading classifier fallback and result auto-evaluation
SkillAFreshness: A★ 711Claude Code +9 - 0.721score
dataforseo
alinaqi
Keyword & SERP research via the DataForSEO API — search volume, competition, CPC, keyword ideas for naming/SEO decisions
SkillBFreshness: B★ 711Claude Code +9 - 0.717score
llm-patterns
alinaqi
AI-first application patterns, LLM testing, prompt management
SkillBFreshness: B★ 711Claude Code +9 - 0.717score
autonomous-testing
alinaqi
AI-driven testing agent that auto-discovers, generates, executes, evaluates, and fixes tests for any project type
SkillBFreshness: B★ 711Claude Code +9 - 0.717score
visual-validation
alinaqi
Autonomous screenshot verification — captures, compares, and AI-evaluates UI changes for visual regressions
SkillBFreshness: B★ 711Claude Code +9 - 0.717score
external-model-delegation
alinaqi
UserPromptSubmit hook pattern that classifies prompts into six cost tiers and delegates to external model CLIs
SkillBFreshness: B★ 711Claude Code +9 - 0.716score
verify-cascade
alinaqi
Verification-gated model escalation — run a cheap model, verify its output with decomposed per-field checks, and escalate to a strong model only when a flag…
SkillBFreshness: B★ 711Claude Code +9 - 0.716score
security-audit
alinaqi
Structured, adversarial, multi-phase security audit of a codebase — recon → coverage-led hunting → finder≠validator validation → machine-readable findings →…
SkillBFreshness: B★ 711Claude Code +9 - 0.713score
context-shunt
alinaqi
Offload large / multi-file reads to a cheap worker model so raw files never enter Claude's context (token savings)
SkillBFreshness: B★ 711Claude Code +9 - 0.708score
ms-teams-apps
alinaqi
Microsoft Teams bots and AI agents - Claude/OpenAI, Adaptive Cards, Graph API
SkillBFreshness: B★ 711Claude Code +9 - 0.708score
ai-models
alinaqi
Latest AI models reference - Claude, OpenAI, Gemini, Eleven Labs, Replicate
SkillBFreshness: B★ 711Claude Code +9 - 0.708score
agentic-development
alinaqi
Build AI agents with Pydantic AI (Python) and Claude SDK (Node.js)
SkillBFreshness: B★ 711Claude Code +9
Signals
- Stars
- 711
- Forks
- 61
- Open issues
- 0
- Last commit (this folder)
- Sep 8, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/alinaqi/maggy
- Website
- www.srooter.ai/maggy
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from alinaqi
- 0.761score
model-routing
alinaqi
9-tier model routing system with cascading classifier fallback and result auto-evaluation
SkillAFreshness: A★ 711Claude Code +9 - 0.721score
dataforseo
alinaqi
Keyword & SERP research via the DataForSEO API — search volume, competition, CPC, keyword ideas for naming/SEO decisions
SkillBFreshness: B★ 711Claude Code +9 - 0.717score
llm-patterns
alinaqi
AI-first application patterns, LLM testing, prompt management
SkillBFreshness: B★ 711Claude Code +9 - 0.717score
autonomous-testing
alinaqi
AI-driven testing agent that auto-discovers, generates, executes, evaluates, and fixes tests for any project type
SkillBFreshness: B★ 711Claude Code +9 - 0.717score
external-model-delegation
alinaqi
UserPromptSubmit hook pattern that classifies prompts into six cost tiers and delegates to external model CLIs
SkillBFreshness: B★ 711Claude Code +9 - 0.717score
visual-validation
alinaqi
Autonomous screenshot verification — captures, compares, and AI-evaluates UI changes for visual regressions
SkillBFreshness: B★ 711Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get build-in-public running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
build-in-public skill. Agent Mart. https://agentmart.dev/skills/alinaqi-build-in-public (last updated 2026-10-24)