straymark-new

Create StrayMark documentation drafts from git context with YAML frontmatter and user confirmation.

2|Updated Apr 2, 2026
One-click install
npx skills add https://github.com/StrangeDaysTech/arborist-cli --skill straymark-new
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: straymark-new
Source: https://github.com/StrangeDaysTech/arborist-cli/tree/main/.gemini/skills/straymark-new
Command: npx skills add https://github.com/StrangeDaysTech/arborist-cli --skill straymark-new

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It turns recent repo changes into the right StrayMark document type and a ready-to-save draft, reducing the manual effort of deciding what to write and where it should live.

Core Features & Use Cases

  • Automatic document type suggestion: Analyzes diffs and code signals to recommend a document type (e.g., AILOG, ADR, ETH, REQ) or uses an explicit type provided by the user.
  • Guided, safe creation flow: Always summarizes the analysis, proposes a filename, and requests confirmation before creating anything.
  • Template- and language-aware generation: Reads .straymark/config.yml to choose English or Spanish templates and fills standardized placeholders (date and sequence).
  • Review-rule enforcement: Applies frontmatter rules that trigger review requirements for high/critical risk and for SEC/MCARD/DPIA documents.

Use case: After a set of commits or a risky refactor, generate an AILOG for complex agent work or an ADR for architectural changes, without guessing the correct type, filename pattern, or folder location.

Quick Start

Run straymark-new with an explicit type to create a draft after analysis (for example: create an AILOG draft based on your latest changes).

Frequently Asked Questions about straymark-new

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate decision records and documentation from git diff analysis?

You can generate decision records from git diff analysis by running a workflow that evaluates recent repository changes to suggest a document type, propose a filename, and fill a standardized template draft. The process requires mandatory user confirmation before writing any files.

What is the best way to automate architecture decision record creation after a refactor?

Automating architecture decision record creation involves analyzing code signals from recent commits to recommend the correct ADR category and generate a ready-to-save draft with deterministic ID sequencing. This removes the manual effort of guessing the correct document type or folder location.

Does StrayMark documentation support language-aware template generation?

Yes, StrayMark documentation supports language-aware template generation by reading configuration files to select either English or Spanish templates. It automatically fills standardized placeholders like date and sequence numbers based on the specified language configuration.

How do review rule requirements apply to high risk repository changes?

Review rule requirements apply to high risk repository changes by enforcing YAML frontmatter rules that trigger mandatory review requirements for high and critical risk levels. These rules also activate for specific security, data protection, and payment card document types.

Can I specify an explicit document type instead of relying on automatic diff analysis?

Yes, you can specify an explicit document type like AILOG or ADR instead of relying on automatic diff analysis. The system will still analyze your latest changes to populate the template draft while respecting your explicitly chosen category and filename pattern.

What limitations exist when creating workflow documentation for repository changes?

A key limitation when creating workflow documentation is that the system requires mandatory user confirmation before creating any files. Additionally, deterministic ID sequencing depends on existing repository context, and template selection requires a valid configuration file.