Skill Publish

Publish an OpenClaw skill to ClawHub with release checks, version metadata, and command generation. Use when you need to prepare a skill for first publicatio...

Security Scan

The skill's requirements and runtime instructions are consistent with a publish-helper: it only needs python3 and the clawhub CLI, runs a local checker script, and prepares a publish command; nothing suggests credential exfiltration or unrelated capabilities—but the shipped helper script content in the prompt appears truncated, so verify file integrity before running.

Purpose & Capability

Name/description match the actual operations: the skill validates a local skill folder, builds a 'clawhub publish' command, and checks metadata. Required binaries (python3, clawhub) are appropriate and proportional. No unrelated environment variables, binaries, or config paths are requested.

Instruction Scope

SKILL.md instructs the assistant to read the target skill directory and run the included local checker script, gather publish inputs, and present the exact 'clawhub publish' command. The doc explicitly requires user confirmation before running the actual publish. Note: executing the generated publish command will interact with the user's local clawhub CLI (and its authentication state). Also, the provided script in the prompt is truncated at the end — verify the on-disk scripts/check_publish_ready.py is complete and review it before running.

Install Mechanism

No install spec is present (instruction-only + included helper script). This is low-risk: nothing is downloaded or installed by the skill itself. The only runtime dependencies are standard: python3 to run the checker and clawhub CLI to confirm login or to publish when requested.

Credentials

The skill requests no environment variables or credentials. It does rely on the user's local clawhub CLI login state (checked via 'clawhub whoami'); that is expected for a publish helper and does not require sharing secrets with the skill package itself.

Persistence & Privilege

always is false and the skill does not request persistent or cross-skill privileges. disable-model-invocation is false (normal). The skill does not modify other skills or system-wide configurations according to provided materials.

Assessment

This skill looks coherent for preparing and validating a ClawHub publish. Before using it: (1) inspect scripts/check_publish_ready.py on disk to ensure it is complete and review its logic (the supplied file appears truncated in the prompt); (2) run the checker locally yourself (python3 ./scripts/check_publish_ready.py /path/to/skill) rather than allowing automated execution if you are unsure; (3) confirm you are logged into clawhub (clawhub whoami) and do not paste or store your tokens inside the skill package; (4) when the assistant shows the generated 'clawhub publish' command, review it carefully for correct path, slug, and metadata before consenting to execute it. If any file is missing or the script is incomplete, do not run it until you obtain a verified copy.

Clawhub Publish

Prepare a skill for ClawHub release, generate the publish command, and verify the release inputs before anything is uploaded.

Quick Start

1. Point this skill at the target skill folder.
2. Run the release checker to catch obvious metadata and packaging issues.
3. Fill in slug, display name, version, changelog, and tags.
4. Review the generated clawhub publish command before running it.
5. Fix any public-facing language, personal path, or placeholder issues before release.

Prerequisites

• python3 available for the local checker script.
• clawhub CLI installed.
• Authentication established with clawhub login or clawhub login --token <token> before actual publish.
• A target skill folder with a valid SKILL.md.

Example Prompts

• Prepare this skill for first publish to ClawHub and generate the exact publish command.
• Check whether this skill is safe to publish, then draft the clawhub publish command for version 1.2.0.
• I need to release a new ClawHub version of this skill. Validate the metadata and show me what to run.
• Audit this skill's release inputs before I publish it to ClawHub.

When to Use This Skill

Use this skill when the goal is to publish or update a skill on ClawHub with fewer release mistakes.

Typical use cases:

• First-time publication of a new skill
• Publishing a new semver version of an existing skill
• Validating slug, name, version, changelog, and tags before release
• Checking whether a local skill package looks publishable

When Not to Use

Do not use this skill for tasks that are primarily about skill quality or listing optimization.

Use a different workflow when you need to:

• Improve search visibility or listing copy: use skill-seo
• Evaluate quality, regressions, or trigger accuracy: use skill-test
• Create a new skill from scratch: use skill-creator
• Debug runtime issues inside the target skill itself

If Publishing Is Inconclusive

If the local checks are clean but release confidence is still low, say what remains unverified.

Common next steps:

• Confirm clawhub whoami before publishing
• Run skill-test for quality concerns
• Run skill-seo for discoverability concerns
• Compare the target skill against the currently published version if this is an update

Pre-Publish Checklist

Before publishing, confirm all of the following:

• SKILL.md has valid frontmatter and no obvious placeholders
• Public-facing copy matches the intended audience language
• No personal file paths, usernames, or internal-only identifiers leak into the package
• Slug, display name, version, changelog, and tags are ready
• Any remaining release risk has been explained to the user

Versioning Guide

Use semantic versioning deliberately:

Change type	Suggested version
First publish	1.0.0
Small wording or typo fix	1.0.1
Content improvement or meaningful documentation update	1.1.0
Major workflow or structure rewrite	2.0.0

If the user already has a versioning policy, follow it instead of this default table.

Workflow

1. Read the target skill's SKILL.md and inspect the directory contents.

2. Run the local checker:

   python3 {baseDir}/scripts/check_publish_ready.py /path/to/skill
   

3. Collect release inputs:

   • slug
   • display name
   • version
   • changelog
   • tags

4. Re-run the checker with release inputs to generate the exact publish command.

5. Present the command and any warnings to the user for confirmation.

6. If the user wants execution, run the generated clawhub publish ... command.

7. After publishing, verify the result with clawhub search "<slug>" or by opening the ClawHub page.

Batch Publish

When publishing multiple skills:

1. Run the local checker for each skill independently.
2. Review slug, version, and changelog for each release.
3. Publish one skill at a time so failures stay attributable.
4. Verify each published slug before moving to the next one.

Commands

# Baseline release audit
python3 {baseDir}/scripts/check_publish_ready.py /path/to/skill

# Release audit plus publish command preview
python3 {baseDir}/scripts/check_publish_ready.py /path/to/skill \
  --slug my-skill \
  --display-name "My Skill" \
  --version 1.0.0 \
  --changelog "Initial release" \
  --tags latest

# Actual publish command after review
clawhub publish /path/to/skill --slug my-skill --name "My Skill" --version 1.0.0 --changelog "Initial release" --tags latest

Definition of Done

• The local checker has run against the target skill.
• Required release inputs have been validated or explicitly flagged as missing.
• A concrete clawhub publish ... command has been prepared.
• Any publish blockers or warnings have been explained to the user.
• If a publish was requested, the post-publish verification step has been completed.

Assistant Responsibilities

• Distinguish publish blockers from optional improvements.
• Do not run clawhub publish until the user confirms the release inputs.
• Be explicit about what was verified locally versus what still depends on ClawHub or authentication.
• Preserve the target skill's actual metadata unless the user asks for edits.

Notes and Constraints

• Local checks cannot guarantee that ClawHub authentication is valid unless the publish command is actually run.
• This skill helps generate and review the release command; it should not silently publish.
• slug should usually match the local directory name unless the user has a clear migration reason.
• Treat version bumps as intentional release actions, not automatic fixes.
• Publishing a technically valid skill is not the same as publishing a good skill; use skill-test and skill-seo if quality or discoverability is still in doubt.

Common Errors

• clawhub CLI missing: install it before attempting publish
• not logged in: run clawhub login, or use clawhub login --token <token>, then confirm with clawhub whoami
• slug rejected or already taken: choose a more specific slug
• invalid semver: correct the version before generating the final command

Resources

• ClawHub publish flow and flags: references/publish-flow.md
• Release checklist and common blockers: references/release-checklist.md
• Local release checklist: references/checklist.md
• Common publish errors and fixes: references/common-errors.md