Guide · Agent skill
How to use the Loading Addressable Markdown agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
Loading Addressable Markdown is an agent skill from the OpenRig project that lets an AI agent read one specific section of a Markdown file, identified by its heading, instead of the whole document. It is for developers and teams whose OpenRig missions, slices, dashboards or tasks point to Markdown sections such as path#h2-slug. After this guide you will know what the skill does, how to install it, how to phrase requests so the agent uses it, and what to check before running it.
Loading Addressable Markdownmvschwarz/openrig · Apache-2.0 Download the skill Quick start
- 1Check Node.js. Confirm Node.js is installed, since the bundled script is an .mjs file run with node.
- 2Download and unzip. Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use or a project's .claude/skills/ folder.
- 3Or upload in the Claude apps. Alternatively, upload the ZIP under Settings → Capabilities → Skills.
- 4Write an address. Reference the section as path#h2-slug or path#h2-slug/h3-slug in your request or mission file.
- 5Test the script. Run the resolver with --root set to your project folder and a quoted address to confirm the output.
- 6Ask the agent. Make a natural request that includes the address, and check that the returned text matches the section you wanted.
What is an agent skill, and what does this one do?
An agent skill is a folder containing instructions (a SKILL.md file) and optionally scripts that an AI agent can use when a task matches the skill's description. This skill's description says it applies when a mission, slice, dashboard or task references Markdown as path#h2-slug or path#h2-slug/h3-slug outside the OpenRig context-pack library.
When the skill is triggered, the agent runs a bundled script, scripts/resolve-markdown.mjs, which returns only the addressed section. The skill's overview describes it as loading exactly one addressed Markdown section from any filesystem tree, so the file does not need to live in the context library.
- Whole file: file.md
- Full H2 section, including its child sections: file.md#h2-slug
- Full H3 section: file.md#h2-slug/h3-slug
How do you install the skill?
Download the ZIP from this page and unzip it. You can install it in one of two ways. For a personal setup, unzip it into ~/.claude/skills/. For a single project, unzip it into that project's .claude/skills/ folder. In the Claude apps you can instead upload the ZIP under Settings → Capabilities → Skills.
The skill's own example command refers to ~/.agents/skills/loading-addressable-markdown/scripts/resolve-markdown.mjs, so check where the script ended up after unzipping and use that path if it differs.
- Personal: ~/.claude/skills/
- Project: .claude/skills/ inside your project
- Claude apps: Settings → Capabilities → Skills, then upload the ZIP
- Requires Node.js, because the script is an .mjs file run with node
How do you trigger the skill with a natural request?
You do not need a special command. Ask for a section by its address, or describe the task in a way that mentions the file and the heading. The skill's trigger is a reference in the form path#h2-slug or path#h2-slug/h3-slug, so including that form in your request makes it more likely to be used.
Slugs are lowercase, Markdown emphasis and code markers are removed, and any other run of non-alphanumeric characters becomes a hyphen. So a heading such as 'Source Cleanup' becomes source-cleanup.
- Read slices/09-source-cleanup/SPEC.md#proposal and summarise it.
- Load the Source Cleanup H3 under the Proposal section of the spec at path#h2-slug/h3-slug form.
- Pull only the proposal section from SPEC.md before you change anything.
How do you run the script directly?
Run the bundled script yourself to test an address. Relative file references resolve from the folder given to --root. Without --root they resolve from the current directory. Quote the address so the shell does not interpret the # character.
Duplicate or missing paths fail loudly rather than returning a guess. Headings inside fenced code blocks are ignored, so a # line inside a code sample will not be mistaken for a section.
- Use --root to set the base folder for relative paths.
- Quote addresses in shell commands.
- Do not use rig context get for arbitrary filesystem paths; that command resolves context-library references.
node ~/.agents/skills/loading-addressable-markdown/scripts/resolve-markdown.mjs \
--root /path/to/mission \
'slices/09-source-cleanup/SPEC.md#proposal'What should you avoid?
Do not slice Markdown by hand with sed or line numbers. The resolver handles span boundaries and code fences, and manual slicing tends to cut sections at the wrong line. Let the agent use the skill, or use the script itself, so the result matches the address exactly.
- Do not hand-slice headings with sed or line numbers.
- Do not expect H1 or H4 and deeper addresses to work; the grammar covers H2 and H3 only.
What are the safety notes?
The skill runs a local script, so it is not just reading instructions. It reads whatever path it is given, so do not point it at sensitive files you would not want a local script to open. The skill does not use the network, according to its listing. It also only reads sections and does not change files.
- Runs scripts: yes (scripts/resolve-markdown.mjs)
- Uses network: no
- Changes files: no, it only reads sections
- Read the script before you rely on it in a sensitive environment
What people use it for
- Loading a specific proposal or spec section from a mission folder for an agent to act on.
- Pulling one H3 subsection from a long planning document without sending the whole file into context.
- Giving an agent exact section text when a dashboard links to a Markdown heading.
- Checking that a referenced slice spec contains the section a task expects.
Frequently asked questions
How do I install the loading-addressable-markdown skill?
Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder. In the Claude apps you can instead upload the ZIP under Settings → Capabilities → Skills.
What do I need to run it?
You need Node.js, because the bundled script scripts/resolve-markdown.mjs is run with node. The skill runs local scripts and does not use the network, according to its listing.
Is the loading-addressable-markdown skill free?
The listing states the licence is Apache-2.0. Check the repository's licence file for the full terms before reuse.
Does it only work with OpenRig files?
No. The skill's overview says it works on any filesystem tree and does not require the file to live in the OpenRig context library.
Why does my section address not work?
Addresses only target H2 and H3 headings, and the path must exist exactly once. Duplicate or missing paths fail loudly. Check that the heading is H2 or H3 and that the slug matches the lowercase, hyphenated form.
Can the skill edit my Markdown files?
No. According to the listing, it only reads sections and does not change files.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
This skill suits OpenRig users and teams whose planning documents are organised with H2 and H3 headings and who need exact section lookups from an agent. It is a poor fit for general document editing, for non-Markdown files, or for documents without H2/H3 structure, and you should avoid it for sensitive files you would not want a local script to read.
Download the skill