DataArt/artisyn-implementation-harness / skills/implementation-harness
implementation-harness
FlaggedInstall the portable AI-dev harness, rebuild its operational graph standalone, harvest project skill improvements into its canonical templates, package direct canonical edits into a portable bundle, or apply canonical updates to an existing project. Use /implementation-harness launch, /implementation-harness reindex-graph, or /implementation-harness patch harvest|prepare|apply.
canonical/skills/design/patterns/docs-writer-authoring.md
View on GitHub# Pattern: Docs-Writer Authoring (Design)
Applies whenever a design-lifecycle skill is about to write or update a design artifact on disk — `design/00 Frame.md`, `design/01 Research*.md`, `design/02 Evaluate*.md`, any `solution/*.md` file, or a content-bearing append to `design/Report.md` (e.g. `patterns/waf-review.md`'s WAF Assessment block). Does not apply to `design/Report.md`'s routine phase-tracker/Efforts-table bookkeeping in the do→report→commit pattern — that's mechanical state tracking, not authored documentation content.
## The rule
Apply `{{provider:skills_root}}/design/runtime/delegation.md`. Delegate authoring to the installed `docs-writer` role when delegation is available and permitted; otherwise use its explicitly recorded inline-authoring fallback. The orchestrating skill still does all the thinking — framing the problem, running the OQ triage, building the tradeoff matrix, deciding the recommendation — but the act of rendering that content to disk as a `.md` file goes through `@docs-writer`, not a direct `Write`/`Edit` call from the orchestrator.
## Why delegate a file the orchestrator already knows how to write
`@docs-writer` enforces conventions the orchestrator does not reliably apply on every write: the certainty principle (resolved facts vs. open questions are never blurred), diagram-first format with correctly `classDef`-styled colored Mermaid (no default-grey diagrams), and DRY across sibling documents. Writing these files inline, phase after phase, drifts from those conventions over a long session. Routing every design-artifact write through one specialist keeps them consistent regardless of which phase or how many phases deep the session is.
## How to apply
1. Assemble the full content for the artifact yourself first — this pattern does not change _what_ goes in the file, only _who_ performs the write. Have the target path, the template to follow (e.g. `{{provider:skills_root}}/design/templates/frame.md`), and the drafted content (problem statement, options, matrix, recommendation, etc.) ready.
2. Delegate to `@docs-writer` with: the target file path, the template file path, the drafted content/decisions to render, and an explicit instruction to apply the certainty principle and diagram-first Mermaid conventions.
For sections destined for architecture documentation, also provide the relevant
Engineering Guide intents and require them to be applied at the target tier:
`docs/Engineering Guide/Documentation/03 HLD Intent.md` for HLD/container
content, `04 LLD Intent.md` for LLD/component content, and `05 NFR Intent.md`
for NFR requirements and acceptance evidence. Read the applicable intent before
drafting; keep HLD, LLD and NFR responsibilities separate, link authoritative
sources, and do not duplicate or weaken their required coverage.
3. Once `@docs-writer` confirms the file is written, proceed to the write-confirm pattern (`{{provider:skills_root}}/design/patterns/write-confirm.md`) exactly as before — the developer still reviews and confirms in their IDE.
## Why
Design sessions produce many artifacts across many phases, each written by whichever phase skill happens to be running. Without a single authoring owner, doc quality and convention adherence (certainty principle, colored Mermaid, DRY) drift phase-to-phase and session-to-session. `@docs-writer` already owns this responsibility for published HLD/LLD content (`design-publish` Step 3) — extending that ownership to every design-lifecycle artifact, not just the final published output, closes the gap where working-draft docs were the one place conventions weren't enforced.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |