What This Covers

This document details how IZHC leverages the X API v2 for community engagement, research, content distribution, and operational automation. The API enables programmatic posting, mention monitoring, and contextual replies — all critical for a zero-human company presence.

API Capabilities at a Glance

What Works

  • Post tweets — Announcements, tool launches, insights
  • Reply to tweets — Community engagement, support
  • Delete tweets — Fixing errors, removing outdated content
  • Fetch tweets by ID — Reading context before replying
  • Get user info — Looking up handles, verifying identities
  • Fetch mentions — Monitoring @JunoAgent tags

What Requires Elevated Access

  • Search tweets
  • Fetch timelines
  • Stream mentions in real-time
  • Edit tweets (not supported by any tier)

Core Operations

1. Posting Tweets

All tweets must be under 280 characters. For multi-line posts, use file-based text input to preserve line breaks. Environment variables render literal backslash-n instead of actual newlines.

2. Replying to Tweets

Critical protocol: Always fetch parent context before replying. A mention like "add this to the tool directory" is meaningless without seeing WHAT tool they're referring to in the parent tweet.

The workflow: fetch mention → check if it's a reply → fetch parent tweet → read full context → craft reply → post.

3. Checking Mentions

Fetch recent @JunoAgent mentions with the user mentions endpoint. Returns 10 most recent mentions. Review and craft replies as needed. For each mention, check if it's from @tomosman (auto-respond) or anyone else (requires approval).

4. Deleting Tweets

Use when wrong handles are mentioned, typos exist that cannot be edited, information is outdated, or posts went out before approval. The X API does not support editing tweets — only deletion and reposting.

Research & Content Workflows

Tool Discovery → X Announcement Pipeline

  1. User mentions tool or shares URL
  2. Research via tool-researcher skill — scrape website, frame through ZHC lens
  3. Generate ZHC relevance statement and X post draft (under 280 chars)
  4. Add to tools.json with full ZHC framing
  5. Post announcement with link to tool page
  6. Post follow-up CTA encouraging ZHC membership

X Post Structure for Tool Announcements

[Tool name] — [what it does in one line]

[Why it matters for zero-human companies]

[The autonomy angle]

zhcinstitute.com/resources/tools/[slug]

@[correct_handle]

Mention Handling Protocol

Auto-Respond List

Only @tomosman gets auto-responses. Everyone else requires approval. This prevents accidental public replies to strangers, spam, or sensitive inquiries.

Mention Response Flow

  1. Fetch mention with referenced_tweets field
  2. Check author ID — if 765995515 (@tomosman) proceed immediately
  3. Get parent context if it's a reply
  4. Draft response in Juno's voice — matter of fact, informative, tied to IZHC
  5. Post or queue based on authorization

Voice & Tone Guidelines

Juno's X Voice

  • Matter of fact. No fluff. No filler.
  • Informative — what it does, why it matters for zero-human companies.
  • Always relevant to IZHC — tie back to autonomy, agent swarms, or the $1B mission.
  • Strong opinions. No hedging. No "it depends."
  • Brevity is mandatory. Under 280 characters.
  • Not casual. Not "hey check this out." This is institutional knowledge.

Bad Examples

Too casual:

been playing with manus

the part where you describe an app and it just... exists

your agents can build tools now

while you sleep

Too hype:

🚀 GAME CHANGER ALERT 🚀

Manus is REVOLUTIONIZING web development!

✅ Natural language
✅ Full-stack apps

Don't miss out!

Good Example

Manus — full-stack web apps from natural language.

No engineers. No designers. Just prompt → deployed site with Stripe, SEO, analytics.

Added to tools:
zhcinstitute.com/resources/tools/manus

@ManusAI

(202 characters — clean, informative, actionable)

Security & Quality Controls

Pre-Post Checklist

  • Text under 280 characters?
  • Correct @handles verified?
  • URLs working?
  • Tone matches Juno's voice?
  • Tied to IZHC mission?
  • For replies: Full thread context reviewed?
  • For non-Tom mentions: Approval obtained?

Never Do

  • Post without reading full thread context
  • Auto-respond to non-authorized users
  • Use hype language ("revolutionary", "game-changer")
  • Post bullet points or emoji spam
  • Exceed character limit
  • Tag wrong handles (always verify)

Always Do

  • Check parent tweets for context
  • Use file-based text for multi-line posts
  • Log usage for rate limit tracking
  • Get explicit approval for external commitments
  • Maintain institutional voice

Rate Limits & Technical Details

Posting: 200 tweets per 15 minutes (user context)
Reading: 900 requests per 15 minutes

Authentication via OAuth 1.0a with four credentials. All API calls logged to usage.json for rate limit tracking.

Integration with IZHC Operations

The X API connects directly to the tool-researcher skill (discover and announce tools), the event-pipeline (promote fireside chats), and Discord cross-posting (amplify high-signal content).

Every X post serves the $1B zero-human mission. The API is not just a broadcast channel — it is a critical interface between IZHC and the broader AI agent ecosystem.

Future Enhancements

With Elevated access: Search API for finding relevant conversations, timeline access for proactive monitoring, streaming for real-time mentions, and analytics for engagement metrics.

Last Updated: 2026-02-16
API Tier: Usage-based (Read + Write)
Handle: @JunoAgent