Skip to content
Back to skills

Lixblogs Publish

ASecurity

Safely preview, publish, unpublish, trash, restore, or permanently delete LixBlogs posts through the supported CLI. Use when an agent is asked to change a post's public or deletion state and must enforce explicit approval, revision checks, and recovery guidance.

  • 85 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added September 3, 2026
devopsbash

Works with

  • cli

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned September 3, 2026

npx -y skills add elixpo/elixpo_chapter --skill lixblogs-publish --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Lixblogs Publish?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Lixblogs Publish
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/elixpo-lixblogs-publish/badge)](https://www.skillsdirectory.com/skills/elixpo-lixblogs-publish)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: lixblogs-publish
description: Safely preview, publish, unpublish, trash, restore, or permanently delete LixBlogs posts through the supported CLI. Use when an agent is asked to change a post's public or deletion state and must enforce explicit approval, revision checks, and recovery guidance.
---

# LixBlogs publish

Use `@elixpo/lixblogs-cli` 1.2.0 or newer with `--json --no-input`. Never use D1, cookies, raw tokens, or direct HTTP calls.

## Access and approval

| Action | Scope | Consequence |
| --- | --- | --- |
| Preview | `lixblogs:blog:read` | No state change |
| Publish / unpublish | `lixblogs:blog:publish` | Adds or removes public access |
| Trash / restore | `lixblogs:blog:delete` | Hides or recovers content |
| Permanent delete | `lixblogs:blog:delete:permanent` | Irreversible removal |

Before a state change, state the target, current state, intended state, and consequence. Obtain explicit user approval. A prior request to draft or edit is not approval to publish or delete. Pass `--yes` only after approval.

## Procedure

1. Verify identity and scopes with `lixblogs whoami --json --no-input`.
2. Inspect the latest state and ETag:

```bash
lixblogs blog preview BLOG_ID --json --no-input
```

3. Preflight the exact operation with `--dry-run`.
4. Show the consequence and ask for approval if it is not already explicit.
5. Execute once with `--yes`, `--etag ETAG`, and a stable `--idempotency-key` where accepted.
6. Report the returned canonical URL and ETag. Fetch once to verify the final state.

```bash
lixblogs blog publish BLOG_ID --etag ETAG --idempotency-key KEY --dry-run --json --no-input
lixblogs blog publish BLOG_ID --etag ETAG --idempotency-key KEY --yes --json --no-input
lixblogs blog publish BLOG_ID --status unlisted --etag ETAG --idempotency-key KEY --yes --json --no-input
lixblogs blog unpublish BLOG_ID --etag ETAG --yes --json --no-input
lixblogs blog trash BLOG_ID --etag ETAG --yes --json --no-input
lixblogs blog restore BLOG_ID --etag ETAG --yes --json --no-input
lixblogs blog delete BLOG_ID --etag ETAG --permanent --yes --json --no-input
```

Trash is the default deletion behavior. Never add `--permanent` unless the user explicitly asks for irreversible deletion after seeing that consequence.

## Conflicts and recovery

- Exit `3` / `revision_conflict`: stop. Fetch the current post, explain that it changed, and request a new decision; never force the stale transition.
- Exit `5` / `confirmation_required`: obtain explicit approval; do not silently retry.
- `idempotency_in_progress`: wait for the original request result before retrying with the same key.
- `idempotency_key_reused`: create a new key only for a genuinely new operation.
- Authentication or missing-scope errors: request only the named scope.
- If the response has a request ID, include it in the failure report. Never print credentials.

Use `lixblogs-author` for content changes before returning to this procedure.

Files in this skill

  • SKILL.md2.9 KB
  • agents/openai.yaml272 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.