← Back to Skills Marketplace
Pandoc
by
Oliver Herklotz
· GitHub ↗
· v1.0.2
· MIT-0
771
Downloads
1
Stars
4
Active Installs
3
Versions
Install in OpenClaw
/install pandoc
Description
Convert documents between formats using pandoc. Supports HTML, Markdown, DOCX, PDF, EPUB, LaTeX, ODT, RST, Org, MediaWiki, JIRA, CSV, Jupyter notebooks, and...
Usage Guidance
This skill is internally consistent and appears benign, but check a few practical things before installing and allowing autonomous use: (1) The package metadata did not declare pandoc as a required binary—ensure pandoc is installed on the agent's host (and install any PDF engine you need) before use. (2) Review scripts/convert.sh (included) yourself — the agent will run pandoc commands exactly as written, so don't pass sensitive files you wouldn't want processed locally. (3) The source/homepage in metadata is 'unknown' / no homepage provided — if you need stronger assurance, prefer a skill that links to an official repo or vendor. (4) If you are uneasy about autonomous invocation, restrict the skill to user-invocable only; otherwise it will run when the agent decides conversion is appropriate.
Capability Analysis
Type: OpenClaw Skill
Name: pandoc
Version: 1.0.2
The skill is a legitimate wrapper for the Pandoc document conversion utility. The helper script `scripts/convert.sh` is well-authored, using bash arrays and proper quoting to safely handle arguments and file paths, which mitigates standard shell injection risks. The documentation in `SKILL.md` and `references/formats.md` provides helpful, context-appropriate instructions for the AI agent without any evidence of malicious prompt injection, data exfiltration, or unauthorized execution logic.
Capability Assessment
Purpose & Capability
The skill's name, description, SKILL.md, and script consistently implement a pandoc-based document converter. However, the registry metadata lists no required binaries while the SKILL.md and scripts clearly require pandoc (and optionally local PDF engines like xelatex, wkhtmltopdf, weasyprint, etc.). The skill should have declared pandoc (and optionally known PDF engines) as required binaries.
Instruction Scope
SKILL.md instructs the agent to use the included scripts/convert.sh and to run pandoc for conversions. The script only operates on user-supplied input/output paths, checks that files exist, validates available local PDF engines, and invokes pandoc. There are no instructions to read unrelated system files, environment variables, or send data to external endpoints.
Install Mechanism
No install spec is present (instruction-only plus a local helper script), so nothing is downloaded or written by an installer. The included bash script is local and executed by the agent; there are no remote download URLs or extract operations in the package.
Credentials
The skill requests no environment variables or credentials, and the script does not attempt to read secrets or external config. It only checks for binaries on PATH and works with files the user provides. This is proportionate to a document-conversion utility.
Persistence & Privilege
The skill does not request permanent/always-on presence (always:false) and does not modify other skills or system-wide agent settings. It runs local commands at invocation time only.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install pandoc - After installation, invoke the skill by name or use
/pandoc - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
- Updates the helper script section to check for input file, output directory, and available PDF engines.
- Troubleshooting guide now covers more PDF engines and includes troubleshooting for CSS not applying to PDF.
- Improved documentation for PDF engine selection and how the helper script handles defaults and errors.
v1.0.1
- Added skill metadata: version, author, and homepage link.
- Introduced a compatibility section listing required and optional PDF engines.
- No changes to usage, commands, or functionality documentation.
v1.0.0
Initial release of the pandoc skill:
- Enables document conversion between a wide range of formats supported by pandoc, including HTML, Markdown, DOCX, PDF, EPUB, LaTeX, ODT, and more.
- Includes a helper script (`scripts/convert.sh`) for easy conversion with auto-detected file formats and sensible defaults.
- Supports custom styling, templates, table of contents, PDF engine selection, and metadata injection during conversion.
- Provides troubleshooting tips for common issues and guidance on best practices for format selection and output customization.
- Offers usage examples for common conversion scenarios and advanced features such as filters, multi-file input, and media extraction.
Metadata
Frequently Asked Questions
What is Pandoc?
Convert documents between formats using pandoc. Supports HTML, Markdown, DOCX, PDF, EPUB, LaTeX, ODT, RST, Org, MediaWiki, JIRA, CSV, Jupyter notebooks, and... It is an AI Agent Skill for Claude Code / OpenClaw, with 771 downloads so far.
How do I install Pandoc?
Run "/install pandoc" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Pandoc free?
Yes, Pandoc is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Pandoc support?
Pandoc is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Pandoc?
It is built and maintained by Oliver Herklotz (@oliver-hrkltz); the current version is v1.0.2.
More Skills