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/agents/product-architect/INTENT.md
View on GitHub---
agent: product-architect
output: {{provider:agents_root}}/product-architect.{{provider:agent_extension}}
---
## Purpose
Read-only governance agent for a shared product-context layer (a docs tree describing the whole product/platform, separate from any single downstream project's implementation). Collaborates at three points in `{{provider:invoke:design}}`: Research (neutral options grounded in platform precedent), an optional Solution review gate (structural completeness + governance-altitude check), and Publish (doc-impact identification + post-write review). This is the deployable form of `{{DESIGN_ARCHITECT_AGENT}}` for a **product-context-scoped** deployment — see the design-lifecycle-scope conditional in `design/INTENT.md`.
**Deploy this agent when the design-lifecycle-scope conditional (see `design/INTENT.md`) resolves to Product-context-scoped _or_ Two-track.** If `{{provider:invoke:design}}` is Project-scoped only (targets this project's own HLD/LLD, no shared product-context layer), skip this agent entirely — use the existing `architect` agent as `{{DESIGN_ARCHITECT_AGENT}}` and do not create `product-architect.md`. In Two-track mode, both this agent and `architect` are deployed side by side — this agent handles the Product track only, `architect` handles the Project track; the deployed `{{provider:invoke:design}}` skill chooses between them per session via `patterns/context-resolution.md`, not at generation time.
## Substitutions
| Placeholder | Source | Notes |
| ------------------------------------- | ------------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `{{PRODUCT_NAME}}` | Q1 — product/platform name | Used in the agent persona line — the product this design work governs, not any single repo's project name |
| `{{DESIGN_DOCS_ROOT}}` | Q10 — design lifecycle scope (see `design/INTENT.md`) | Root path of the shared product-context docs tree this agent is custodian of: `docs/{{PRODUCT_CONTEXT_DIR}}/`. **Always the fixed Product-track path here, even in Two-track mode** — this agent's identity is the Product track, so it never uses the `{source docs}`/`{dest docs}` runtime tokens that other `{{provider:invoke:design}}` skills use in Two-track mode |
| `{{DOWNSTREAM_PROJECTS_NOTE}}` | Q — downstream project topology | One or two sentences naming the downstream projects/repos that implement the governance layer's principles — e.g. "The N platform projects — X, Y, Z — plus shared Infrastructure, each implement these principles in their own repositories (repo-a, repo-b, repo-c)." Omit repo names entirely if they're not publicly nameable in the agent body. |
| `{{PRODUCT_CONTEXT_GOVERNANCE_PATH}}` | Ask at bootstrap time (or infer from Stage 1c if the docs tree already exists) | The specific subfolder (and file, if there's a canonical overview doc) within `{{DESIGN_DOCS_ROOT}}` that holds the governance principles — **not assumed to be named `Architecture/`**. Ask: "Within `{{DESIGN_DOCS_ROOT}}`, which subfolder holds the governance principles / architectural boundaries every downstream project must honor?" Suggest `{{DESIGN_DOCS_ROOT}}Architecture/` as a default only if that subfolder actually exists or the developer has no preference — do not assume it. Resolves to a path like `docs/Product Context/Architecture/` or `docs/Product Context/Architecture/00-overview.md` depending on whether a canonical overview file exists. |
## Conditionals
- If this deployment has only one repository (no separate downstream implementation repos) and is Project-scoped: this agent doesn't apply — skip it, use `architect` instead.
- `{{PRODUCT_CONTEXT_GOVERNANCE_PATH}}` is deployment-specific — never hardcode a subfolder name (e.g. `Architecture/`) as a fallback inside the deployed agent body. If unresolved at generation time, ask the developer rather than guess.
## Verify checklist
- [ ] `{{PRODUCT_NAME}}`, `{{DESIGN_DOCS_ROOT}}`, and `{{PRODUCT_CONTEXT_GOVERNANCE_PATH}}` are not left as literal placeholders in the deployed agent
- [ ] No hardcoded `Architecture/` subfolder name — the deployed agent's governance-path reference matches this project's actual docs structure
- [ ] `{{DESIGN_DOCS_ROOT}}` matches `docs/{{PRODUCT_CONTEXT_DIR}}/` — the fixed Product-track path, never a runtime token, even when deployed alongside Two-track `{{provider:invoke:design}}` skills
- [ ] Agent never claims write access — every write is delegated to `@docs-writer`
- [ ] "Governance, not implementation" framing preserved — this agent must refuse to prescribe downstream implementation detail
- [ ] `tools:` stays `Read, Glob, Grep` — no `Write`/`Edit`/`Bash`, this agent is read-only by design
- [ ] Deployed when the design-lifecycle-scope conditional is Product-context-scoped or Two-track (see `design/INTENT.md`) — not deployed alongside a Project-scoped-only `{{provider:invoke:design}}`
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 |