Skip to content
SkillModel-led1,408 tokens in use

document-app

Writes or refreshes an app's system docs from its code (architecture, flows, permissions, variables), or checks existing docs against the code and reports disagreements.

680 stars on the repository MIT Licensephuryn/product-engineering · ai-shipping/skills/document-app
claude plugin install ai-shipping@product-engineering
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing document-app costs

39tokens idle
1,408tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

Nothing detected: no shell, network, file writes, git or credentials in its files.

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Excellent
  • Frontmatter8/15
  • Structure25/25
  • Examples9/15
  • Output15/15
  • Size15/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Good
  • Development activity100
  • Maintenance100
  • Community44
  • Maturity4
  • Momentum40

document-app reverse-engineers the durable documentation an AI-built app is usually missing. It reads the code as the source of truth and produces core documents for architecture, permission-relevant flows, roles and permissions, and config and secrets, plus conditional documents for emails, cron, SEO and embedded agents when those capabilities exist.

Those documents become the baseline that later audits compare the code against. A check-only mode reads existing docs and code, prints every factual disagreement with both sides cited, and writes no file.

What You Get

  • architecture.md with the stack, auth flow and trust boundaries, linking to each other document produced
  • flows.md listing permission-relevant journeys, each protected step’s authorization check and the side effects it causes
  • permissions.md with roles, scope derivation and a resource by operation by role matrix, noting which checks are row-level security and which are code
  • variables.md mapping config and secrets to risk and rotation, plus conditional emails.md, cron.md, seo.md and automation.md where they apply
  • A check-only report of disagreements, behaviors present in code but in no doc, and what was not checked

Common Use Cases

  • Giving an AI-built app the written intent that a security or performance audit needs
  • Refreshing docs before a release and seeing where they drifted from the code
  • Handing a reviewer a map of who can do what and where the risk lives
  • Checking hand-maintained docs against the code without letting an agent overwrite them

What Makes It Unique

It treats the codebase as untrusted input to describe, so instructions embedded in code are never followed. It follows the repo’s own docs location and names first and uses documentation/ only where none exist. For large scopes it gives parallel subagents one core document each. Check mode leaves the choice of which side is wrong to the owner.

Under The Hood

The skill is a single markdown file. The agent reads source files in the repository, may start parallel subagents for large scopes, and writes markdown documents into the docs folder. Check mode only reads and prints. It runs no scripts and calls no web services, and it relies on the shipping-artifacts skill for the contents of each document.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers