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/dev/interfaces/intent.md
View on GitHub# Intent interface
`0-intent.md` is the first document of a ticket. Context Gathering writes it once,
before any branch or plan work, so what reaches planning is what the originator
meant and not something several steps removed from it. It follows the capture-intent
idea from the AI-native SDLC playbook: human readable, version controlled, in the
originator's own terms, no formal language required.
This is an interface, not a template. It says what the document must answer. It
does not prescribe headings, order, length or diagrams. Use whatever shape makes the
answers clearest for this ticket; a short ticket may be ten lines.
## What it must answer
- **What is the problem?** What people cannot do today, or what is wrong, and why
it matters. In the words of the ticket, the design or the developer, not
analyst-speak.
- **What does better look like?** The outcome from the user's or developer's point
of view. What "done" means, not how it is built.
- **Who and what is affected?** The users, teams, apps and HLD container(s) that
feel this change.
- **What is out of scope?** Explicit boundaries the developer or ticket set.
- **What are the constraints?** Things that must stay unchanged, deadlines,
compliance or NFR limits, an approved design that is the spec.
- **What is still open?** Questions nobody could answer during Context Gathering.
Empty is fine; never invent one.
Also keep the provenance at the top: ticket, date, and where the intent came from
(tracker, Confluence pages, approved design artifacts, the developer's message),
with links. A reader should be able to check every claim against its source.
## What it must not contain
- Implementation steps, file lists or commands. That is `1-plan.md`.
- Acceptance criteria as a checklist. The plan derives them from this document.
- The chosen solution or a diagram of it. That is the plan's `## Solution`.
- Status, progress or findings. That is `2-report.md`.
## Rules
- Write it in Context Step 4; the developer signs it off in Step 5. Afterwards only
append to the open questions; never rewrite the confirmed answers.
- Developer input wins over tracker text when they conflict; say so where it happens.
- If an approved `solution/Architecture.md` exists, say that it is the spec and
link it. Do not copy it.
- Keep it short. If it is longer than the plan, something that belongs in the plan
has leaked in.
## What good looks like
A developer who has never seen the ticket reads `0-intent.md` and can say in their
own words what is being asked, for whom, where the edges are, and what is unknown.
They cannot tell from it how it will be built, and they do not need to.
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 |