---

name: Blog Management description: Managing Cyril's Workshop blog posts and devblog via API - creating, editing, publishing, and updating content programmatically

Blog Management

This skill enables programmatic management of blog posts on steponnopets.net, including both the main Cyril blog and the devblog. Use this skill when the user asks to edit, update, or fix published blog posts.

Instructions

When the user requests blog post edits or updates, follow these guidelines:

1. Identify the Target Post: Determine which post to edit from the user's request (by slug, "latest", or content description)

2. Read API Key: Always read the API key from ~/.claude/cyril-api-key before making API calls

3. Fetch Current Content: GET the post to see current state before making changes

4. Make Targeted Updates: Only PATCH the fields that need to change (title, content, tags, or publish status)

5. Confirm Success: Report the update with the post URL so the user can verify

6. Handle Both Blogs: DevBlog is at /devblog/api, main Cyril blog is at /cyril/api

Examples

Example 1: Fix a Typo

User: Fix the typo in the admin panel post, change 'teh' to 'the'

Claude: I'll fetch the post, fix the typo, and update it:
- Read API key from ~/.claude/cyril-api-key
- Fetch post content from /devblog/api/posts
- Find the post about admin panel
- Replace 'teh' with 'the' in the content
- PATCH the updated content back

[Uses the PATCH Update pattern from reference material below]

Updated post "admin-panel-and-tag-filtering": https://steponnopets.net/devblog/#/post/admin-panel-and-tag-filtering

Example 2: Update a Link

User: In the latest devblog post, change the repo link to https://github.com/lawless-m/Robocyril

Claude: I'll update the GitHub link in the most recent post:
- Fetch all posts from /devblog/api/posts
- Identify the latest by published_at timestamp
- Update the GitHub URL in the content
- PATCH just the content field

[Uses the PATCH Update pattern from reference material below]

Example 3: Add Tags

User: Add 'automation' and 'admin' tags to the post about the admin panel

Claude: I'll update the tags for that post:
- Find the admin panel post by searching titles/slugs
- Get current tags array
- Add 'automation' and 'admin' to the array
- PATCH just the tags field

[Uses the PATCH Update pattern from reference material below]

Example 4: Change Title

User: Change the title of the post 'old-slug' to 'Better New Title'

Claude: I'll update the title while keeping the same slug:
- Fetch post by slug 'old-slug'
- PATCH with new title only
- Note: Slug stays the same (it's the identifier)

[Uses the PATCH Update pattern from reference material below]

---

Reference Implementation Details

The sections below contain working patterns and exact API formats for managing blog posts.

API Endpoints

DevBlog

• Base URL: https://steponnopets.net/devblog
• Get all posts: GET /devblog/api/posts
• Get with drafts: GET /devblog/api/posts?include_drafts=true (requires auth)
• Get single post: GET /devblog/api/post?slug=<slug>
• Create post: POST /devblog/api/posts
• Update post: PATCH /devblog/api/post?slug=<slug>
• Delete post: DELETE /devblog/api/post?slug=<slug>

Cyril's Workshop (Main Blog)

• Base URL: https://steponnopets.net/cyril
• Same API structure as DevBlog, just replace /devblog/ with /cyril/

Authentication Pattern

Location: ~/.claude/cyril-api-key Purpose: Authenticate all write operations

All POST, PATCH, and DELETE requests require:

-H "X-Cyril-Key: $(cat ~/.claude/cyril-api-key)"

Key Points:

• API key is stored as single line with no newline
• Same key works for both Cyril blog and DevBlog
• GET requests for published posts don't need auth
• GET with ?include_drafts=true requires auth

GET Posts Pattern

Purpose: Fetch current post content before editing

Get all published posts:

curl https://steponnopets.net/devblog/api/posts

Get single post by slug:

curl "https://steponnopets.net/devblog/api/post?slug=my-post-slug"

Response format:

{
  "slug": "post-title",
  "title": "Post Title",
  "content": "# Markdown content...",
  "tags": ["rust", "web"],
  "repo": "https://github.com/user/repo",
  "created_at": "2025-01-15T10:30:00Z",
  "updated_at": "2025-01-15T10:30:00Z",
  "published_at": "2025-01-15T10:30:00Z"
}

Key Points:

• Drafts have published_at = null
• Tags are JSON array
• Slugs are auto-generated from titles (lowercase, hyphens)
• Slugs cannot be changed after creation

PATCH Update Pattern

Purpose: Update specific fields of an existing post

ALWAYS use this exact pattern:

curl -X PATCH "https://steponnopets.net/devblog/api/post?slug=<slug>" \
  -H "Content-Type: application/json" \
  -H "X-Cyril-Key: $(cat ~/.claude/cyril-api-key)" \
  -d '{
    "content": "# Updated content..."
  }'

All fields are optional - only include what you want to update:

Field	Type	Description
title	string	Update post title (slug stays same)
content	string	Update full markdown content
tags	array	Replace all tags (["tag1", "tag2"])
publish	boolean	true to publish, false to unpublish

Example updating multiple fields:

curl -X PATCH "https://steponnopets.net/devblog/api/post?slug=my-post" \
  -H "Content-Type: application/json" \
  -H "X-Cyril-Key: $(cat ~/.claude/cyril-api-key)" \
  -d '{
    "title": "New Title",
    "tags": ["rust", "svelte", "automation"],
    "publish": true
  }'

Key Points:

• Only send fields you want to change
• Slug in URL identifies the post
• Slug cannot be modified
• Returns updated post on success

POST Create Pattern

Purpose: Create a new blog post (usually done via /blog command)

IMPORTANT: Always use the file-based approach to avoid JSON escaping issues, especially on Windows.

Format (recommended - works on all platforms):

# 1. Write post data to temporary file
cat > /tmp/post.json << 'EOF'
{
  "title": "Post Title",
  "content": "# Markdown content here...",
  "repo": "https://github.com/user/repo",
  "tags": ["® RepoName", "rust", "web"],
  "publish": true
}
EOF

# 2. POST the file
curl -X POST https://steponnopets.net/cyril/api/posts \
  -H "Content-Type: application/json" \
  -H "X-Cyril-Key: $(cat ~/.claude/cyril-api-key)" \
  --data-binary @/tmp/post.json

# 3. Clean up
rm /tmp/post.json

Windows alternative:

# Use Write tool to create post.json in current directory
# Then:
curl -X POST https://steponnopets.net/cyril/api/posts `
  -H "Content-Type: application/json" `
  -H "X-Cyril-Key: <api-key>" `
  --data-binary @post.json

Response:

{
  "slug": "post-title",
  "url": "https://steponnopets.net/cyril/#/post/post-title"
}

Key Points:

• For GitHub repos, first tag MUST be project tag: ® RepoName
• Project tag auto-creates/updates project on Projects page
• Set "publish": false to save as draft
• Slug is auto-generated from title
• ALWAYS use --data-binary @file.json to avoid shell escaping issues

DELETE Pattern

Purpose: Delete a blog post

curl -X DELETE "https://steponnopets.net/devblog/api/post?slug=my-post-slug" \
  -H "X-Cyril-Key: $(cat ~/.claude/cyril-api-key)"

Key Points:

• Requires authentication
• Permanent deletion
• No undo available

Database Schema

Posts are stored in SQLite at /var/www/data/devblog.db and /var/www/data/cyril.db:

CREATE TABLE posts (
  slug TEXT PRIMARY KEY,
  title TEXT NOT NULL,
  content TEXT NOT NULL,
  tags TEXT,  -- JSON array: ["tag1", "tag2"]
  repo TEXT,
  created_at TEXT,
  updated_at TEXT,
  published_at TEXT  -- NULL for drafts
)

Project Tags System

Purpose: Auto-create project entries from blog posts

When a post has a project tag (® ProjectName) as the first tag:

• A project entry is auto-created/updated in the database
• The project description is extracted from the first paragraph of the post
• The project links to the GitHub repo specified in repo field
• Clicking the project tag shows all posts for that project

Example:

{
  "title": "Building Robocyril Admin Panel",
  "tags": ["® Robocyril", "rust", "svelte"],
  "repo": "https://github.com/lawless-m/Robocyril",
  "content": "This post describes building an admin panel.\n\nMore details..."
}

Creates/updates project:

• Name: Robocyril
• Description: "This post describes building an admin panel."
• URL: https://github.com/lawless-m/Robocyril

Common Request Patterns

When the user makes these requests, use these patterns:

User Request	Action	Fields to PATCH
"Fix typo X to Y"	Search & replace in content	content
"Add tag Z"	Append to tags array	tags
"Change title to X"	Replace title	title
"Update GitHub link"	Replace URL in content	content
"Unpublish post X"	Set publish to false	publish
"Publish draft X"	Set publish to true	publish

Error Handling

Status	Meaning	Action
404	Post not found	Verify slug, check if post exists
401	Invalid API key	Check ~/.claude/cyril-api-key contents
400	Invalid JSON	Verify JSON syntax in request
500	Server error	Check vsprod logs

Admin Panel Reference

Web-based admin available at https://steponnopets.net/devblog/#/admin:

• View all posts (including drafts)
• Edit post content, title, tags in modal editor
• Publish/unpublish posts
• Delete posts
• Manage projects
• Requires API key authentication

Best Practices

1. Always fetch before updating - GET the current content first to avoid overwriting unintended fields
2. PATCH only what changed - Don't send all fields, just the ones being updated
3. Preserve existing data - When updating tags, include existing tags plus new ones
4. Confirm with URL - Always return the post URL after updating so user can verify
5. Handle drafts carefully - Check published_at to distinguish drafts from published posts