# Reverse Document

> Create technical documentation from existing implementation using the shared canonical templates.

- **Type:** Skill
- **Install:** `agentstack add skill-avelrl-skills-reverse-document`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [avelrl](https://agentstack.voostack.com/s/avelrl)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [avelrl](https://github.com/avelrl)
- **Source:** https://github.com/avelrl/skills/tree/main/core/reverse-document

## Install

```sh
agentstack add skill-avelrl-skills-reverse-document
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

Purpose: Recover shared technical documentation from existing implementation without inventing intent.

Use when:
- code exists first and a technical design or architecture doc is missing
- you need a defendable document based on implementation facts

Do not use for:
- speculative design before implementation
- game-system reverse docs that belong to `gamedev/`
- user-story or product requirement writing

Inputs / Required Context:
- required: `` and a real code or module path
- read: target path plus nearby code, tests, configs, and relevant docs
- ask only for intent that cannot be recovered from implementation

Outputs / Owned Artifacts:
- owns `docs/design/[slug].md` for `technical-design`
- owns `docs/architecture/adr-[NNNN]-[slug].md` for `architecture`
- creates `docs/design/` or `docs/architecture/` if needed
- uses shared templates only

Modes or Arguments:
- `technical-design|design `
- `architecture `

Execution Rules:
1. Validate the type and target path.
2. Read only the material needed to document responsibilities, interfaces, data flow, dependencies, edge cases, and constraints.
3. Add a short reverse-documentation context near the top of the doc.
4. For architecture outputs, determine the next ADR number before saving.
5. Save the result to the canonical shared path.

Failure / Stop Conditions:
- stop if the path does not exist or the type is unsupported
- stop and redirect if the request is actually a game-system design doc
- if rationale cannot be proven, mark it as inferred or open instead of pretending certainty

Return Format:
- source path
- output path
- what is now documented
- what is still inferred or unresolved
- what should be reviewed next

Example Invocation:
- `/reverse-document technical-design src/auth/refresh.go`

Related Skills / Boundary:
- use `architecture-decision` when the main task is making or updating an ADR
- do not pull game-specific templates or rules into this shared skill

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [avelrl](https://github.com/avelrl)
- **Source:** [avelrl/skills](https://github.com/avelrl/skills)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-avelrl-skills-reverse-document
- Seller: https://agentstack.voostack.com/s/avelrl
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.