Create deep technical dives with contrarian attitude and annotated code for skeptical senior developers. Production reality over polished demos.

Prerequisites

Load the Session Playbook first. Verify your session and profile before starting.

IMPORTANT: NEVER start, stop, or restart services. The API is already running.

CRITICAL: Context Management

New content = New context. ALWAYS. See the Contexts Playbook for full context management commands.

WRONG: Reusing hackernoon-general context for multiple articles RIGHT: hackernoon-agent-mesh-critique, hackernoon-rag-postmortem, hackernoon-llm-benchmarks (one context per article)

CRITICAL: One Step = One Message

Step Action Separate Message?
Step 3 Plan YES
Step 4 Research YES
Step 5 Create YES
Step 6 Image YES

Step 1: Create a FRESH Context

// MCP: systemprompt_cli
{ "command": "core contexts new --name \"hackernoon-[topic-slug]\"" }

// MCP: systemprompt_cli
{ "command": "analytics content stats --since 30d" }

Step 3: Plan the Technical Deep Dive

HackerNoon readers expect code, data, and contrarian takes.

Step 4: Research

// MCP: systemprompt_cli
{ "command": "admin agents message hackernoon -m \"Research this topic.\" --context-id \"$CONTEXT_ID\" --blocking --timeout 120" }

Step 5: Create Content

// MCP: systemprompt_cli
{ "command": "admin agents message hackernoon -m \"Create the article with annotated code examples.\" --context-id \"$CONTEXT_ID\" --blocking --timeout 300" }

Step 6: Generate Image

// MCP: systemprompt_cli
{ "command": "admin agents message hackernoon -m \"Generate a featured image.\" --context-id \"$CONTEXT_ID\" --blocking --timeout 60" }

Step 7: Publish and Verify

// MCP: systemprompt_cli
{ "command": "infra jobs run blog_image_optimization" }
{ "command": "infra jobs run publish_content" }
{ "command": "cloud sync local content --direction to-disk --source hackernoon -y" }

Step 8: Update AI Provenance (MANDATORY AFTER CRUD)

CRITICAL: After ANY content CRUD operation (Create, Read, Update, Delete), agents MUST update the AI Provenance metadata.

AI Provenance is rendered in the content template as a transparency block showing readers how the content was created.

8.1 Update Provenance via CLI

// MCP: systemprompt_cli - Set provenance fields
{ "command": "core content edit [slug] --source hackernoon --set agent=hackernoon --set agent_summary=\"Created article: [Title] - [Why]\" --set category=\"Technical Deep Dive\"" }

Verify provenance is set:

// MCP: systemprompt_cli
{ "command": "core content show [slug] --source hackernoon" }

8.2 Agent Summary Guidelines

The agent_summary is displayed as "Why This Was Created". It should:

  • Be concise (1-2 sentences)
  • Explain the PURPOSE, not just the action
  • Include the content title
  • Reference the contrarian angle if relevant

GOOD: "Created article: Agent Mesh Architecture Critique - challenging single-agent patterns with production data." BAD: "Created article" (too vague)

Provenance is NOT optional. All AI-generated content MUST have complete provenance metadata.

Content Types

Type Length Use Case
hackernoon_article 2,000-4,000 words Deep technical dives

Format Requirements

  • Annotated code blocks (minimum 3)
  • Production metrics and data
  • Contrarian take clearly stated
  • British English

Troubleshooting

Issue Solution
Content not created Verify with core content search
Agent says "created" but doesn't exist Create a NEW context