Expose a queryable graph of linked Markdown plans with YAML frontmatter.
npx skills add https://github.com/sultanvaliyev/anchormd --skill anchormd
Please help me install this Agent Skill. Skill: anchormd Source: https://github.com/sultanvaliyev/anchormd/tree/main/skill Command: npx skills add https://github.com/sultanvaliyev/anchormd --skill anchormd
AnchorMD provides a persistent, queryable project context for AI coding agents by exposing a graph of linked Markdown plans and their relationships to help them understand the codebase.
Run anchormd context to load the project overview and then use anchormd find to locate relevant plans.
High-intent search queries and answers about installing and using this skill.
You can maintain persistent project context for AI by using linked Markdown plans that form a queryable graph. This exposes relationships between plans, helping AI agents understand the codebase across interconnected projects.
The best way to search Markdown plans is using full-text and semantic hybrid search across the linked graph. This surfaces relevant project context and allows you to jump to specific sections of the plans.
You update AI project context after implementing tasks by using simple write and read workflows. This ensures the linked Markdown plans reflect the latest codebase changes and maintain accurate context.
A YAML frontmatter schema helps AI agents discover project plans by defining the name and description metadata. This schema drives the operational logic embedded in the Markdown body to execute behavior.
Yes, you can use linked Markdown plans to navigate large interconnected projects. They provide a centralized overview of plans and their relationships, enabling AI agents to jump to specific sections and query relevant context.
AI coding agents lose context in large projects because they lack a persistent overview of plans and relationships. Exposing a queryable graph of linked Markdown plans solves this by providing the needed structural context.
mecolejrPlan shot-by-shot storyboards for TikTok, Reels, YouTube Shorts, and UGC campaigns.
RigohlCoordinate hybrid semantic and full-text search across multiple engines via FFI.
LuminaVaultGenerate and save markdown development plans to the .hermes/plans/ directory.
skogaiMap codebase relationships between functions, classes, and modules using knowledge graph tools.
