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/context-resolution.md
View on GitHub---
name: Context Resolution
description: Resolves which track (Product / Project) a design session belongs to. Applied exactly once, in Context Gathering Step 3. Every other phase reads the result from design/Report.md's Context section instead of re-resolving it — no branching logic elsewhere in the lifecycle.
type: pattern
---
> ⚠️ **Two-track deployments only.** Do not deploy this file for Project-scoped or Product-context-scoped deployments — this branch is selected during installation; single-track installs omit it.
Resolved once, in Context Gathering Step 3 — this is the only place these
values are decided; no later phase re-derives them.
This is an intelligent-approval checkpoint (`patterns/approval-intelligent.md`).
Auto-approval condition: the ticket, linked docs, or developer's own words
name one track unambiguously — e.g. they reference a specific
`{{PRODUCT_CONTEXT_DIR}}` doc or product-context terminology (Product), or
they reference this repo's own containers, HLD, or NFR docs by name (Project)
— and nothing in the input points the other way.
Condition met → auto-approve: state the resolved track and why in one line
(the specific reference that made it unambiguous), and continue.
Condition fails or is uncertain (nothing names a track, or the input could
plausibly be either) → this is a sign-off gate (`patterns/approval-sign-off.md`):
> "Is this a **Product** design (cross-project product architecture — lands in `docs/{{PRODUCT_CONTEXT_DIR}}/`) or a **Project** design (this repo's own target architecture — lands in `docs/Target Architecture (HLD)/`)? {If the input leans one way: "My read: probably {track}, because {reason} — confirm or correct."}"
**Stop and wait.**
Either way, look up the resolved track in the table below.
| Field | Product | Project |
| -------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------- |
| Source docs | `docs/{{PRODUCT_CONTEXT_DIR}}/` | `docs/Target Architecture (HLD)/` |
| Dest docs | `docs/{{PRODUCT_CONTEXT_DIR}}/{subfolder}` | `docs/Target Architecture (HLD)/{subfolder}` |
| Stories root | `docs/Stories/product-design/` | `docs/Stories/design/` |
| Architect agent (research/review) | `@product-architect` | `@architect` |
| Writer agent | `@docs-writer` | `@docs-writer` |
| Implementation note (Publish hand-off) | Most Product designs are implemented in the owning downstream platform repo(s) — offer `{{provider:invoke:dev}}` in this repo only if the implementation tickets genuinely target this repo's own code. | Project designs cover this repo's own containers — `{{provider:invoke:dev}}` in this repo is the expected next step. |
**Subfolder is not enumerated here** — it can be any subfolder/section under the resolved Source/Dest docs path, at any depth (some product-context layers keep subfolders one level down; some target-architecture trees nest them under a version folder, e.g. `docs/Target Architecture (HLD)/v1/Containers/`). Glob recursively under the resolved Source docs path for existing subdirectories, list what exists, and ask the developer which one this design belongs to (or a new name if none fit).
**Stories root separates where the two tracks physically live on disk**, same idea as Source/Dest docs but for `docs/Stories/`: a Product design's story folder, archive, and every path reference live under `docs/Stories/product-design/`; a Project design's live under `docs/Stories/design/`. The two never mix in the same parent folder, so a reader scanning `docs/Stories/` can tell which track a story belongs to without opening it.
Write the resolved Track, Source docs, Dest docs, Stories root, Subfolder, Architect agent, and Implementation note into the `## Context` section of `design/Report.md`. Every subsequent phase and satellite skill reads them from there — as `{resolved architect agent}`, `{source docs}` / `{dest docs}` / `{stories root}`.
**Then read the resolved Source docs path.** Once Track and Subfolder are
settled, read the relevant docs under the resolved Source docs path for
anything already established on this topic, and record each one in
`design/Report.md`'s `## Evidence` table (`Source` / `Read at stage` =
`context-gathering` / `Relevant decision or constraint` / `Reference/version`).
This is the lifecycle's one context-reading pass for this ticket — Frame,
Research, and every later phase read this table instead of re-reading the
source docs themselves, and it's how session continuity recaps context after
a new session picks the ticket back up. Which path this reads from is itself
track-dependent, so the instruction lives here with the rest of the
track-dependent resolution, not as a separately templated sentence elsewhere.
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 |