Skip to content

Agent Skill: Document Review

MRSF includes a ready-to-use Agent Skill that teaches AI agents to review Markdown documents using the MCP server.

What is an Agent Skill?

Agent Skills are a portable, open format for giving AI agents new capabilities. A skill is a folder with a SKILL.md file containing structured instructions that any skills-compatible agent (Claude Code, Copilot, Junie, OpenHands, etc.) can discover and use.

The mrsf-review Skill

The skill lives at examples/mrsf-review/SKILL.md and instructs an agent to:

  1. Discover the sidecar for the target document
  2. Check existing comments for context
  3. Read and review the document, adding anchored comments with type and severity
  4. Validate the sidecar after adding comments
  5. Summarize findings

Each comment is anchored to a specific line with a category (accuracy, clarity, suggestion, style, issue, question) and severity (low, medium, high).

If an agent needs to attach tool-specific metadata, the MCP tools also support an explicit extensions object whose keys must start with x_. Those entries are stored on disk as flat x_* fields on the comment.

Installation

Copy the skill folder into your project:

bash
cp -r examples/mrsf-review .agent/skills/

Or clone directly:

bash
mkdir -p .agent/skills
curl -sL https://raw.githubusercontent.com/wictorwilen/MRSF/main/examples/mrsf-review/SKILL.md \
  -o .agent/skills/mrsf-review/SKILL.md --create-dirs

Prerequisites

The MRSF MCP server must be configured for your agent. See MCP Server for setup instructions.

Full Skill File

Below is the complete SKILL.md — copy it as-is into your project:

MRSF Document Review

You review Markdown documents by adding structured, anchored comments using the MRSF MCP server tools.

Setup

The MRSF MCP server must be available. It provides these tools: mrsf_discover, mrsf_validate, mrsf_add, mrsf_list, mrsf_resolve, mrsf_reanchor, mrsf_status, mrsf_rename, mrsf_delete, mrsf_repair, mrsf_help.

Workflow

  1. Discover the sidecar for the target document using mrsf_discover.
  2. Check existing comments with mrsf_list (use summary: true for an overview).
  3. Read the document, then add comments with mrsf_add for each issue found. Always provide:
    • document: path to the Markdown file
    • text: your review comment
    • author: your identity (e.g. "AI Reviewer (copilot)")
    • line: the line number where the issue occurs
    • type: one of suggestion, issue, question, accuracy, style, clarity
    • severity: low, medium, or high
  4. Validate the sidecar with mrsf_validate after adding comments.
  5. Summarize what you found using mrsf_list with summary: true.

Comment Guidelines

  • Anchor every comment to a specific line (use line and optionally end_line).
  • Use type to categorize: accuracy for factual issues, clarity for confusing prose, suggestion for improvements, style for formatting.
  • Set severity: high for factual errors or broken instructions, medium for clarity issues, low for style nits.
  • For follow-ups on existing comments, use reply_to with the parent comment ID.
  • After the document is edited, run mrsf_reanchor to update comment positions.

Released under the MIT License.