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/scrum-master/INTENT.md
View on GitHub---
skill: agents/scrum-master
output: {{provider:agents_root}}/scrum-master.{{provider:agent_extension}}
---
## Purpose
Jira workflow agent that centralises all issue reads, transitions, assignments, and ticket creation for the dev and design lifecycles. Skills delegate all Jira operations here instead of calling MCP tools inline — keeps Jira logic in one place and makes it easy to adapt when transition IDs or project keys change.
## Substitutions
| Placeholder | Source | Notes |
| -------------------------------------- | ------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `{{PROJECT_KEY}}` | Q1 — Jira project key | e.g. `SP`, `PROJ`. Used in issue keys, error messages, and operation examples |
| `{{JIRA_URL}}` | Q8 — Jira base URL | e.g. `https://myorg.atlassian.net` |
| `{{CONFLUENCE_URL}}` | Q8 — Confluence base URL | e.g. `https://myorg.atlassian.net/wiki` |
| `tools:` MCP server name (rule) | Q5 — tracker MCP server | Rendering rule, not a literal token: the `tools:` line ships as `mcp__atlassian__*`; rewrite `atlassian` to the project's actual Jira MCP server name, keeping the `mcp__<server>__*` shape (the Copilot renderer derives `<server>/*` from it and rejects anything else). Organisations with an existing connection keep their own server name |
| `{{JIRA_STATE_IN_PROGRESS}}` | Q5 — Jira workflow | The project's own state name for "work started", e.g. `In Progress`, `In Development`. The generic name stays in the first column; this is what the transition actually targets |
| `{{JIRA_STATE_TESTING}}` | Q5 — Jira workflow | Project state for "PR submitted / in review", e.g. `Testing`, `In Review`, `Code Review` |
| `{{JIRA_STATE_DONE}}` | Q5 — Jira workflow | Project state for "complete", e.g. `Done`, `Ready For Release`, `Closed` |
| `{{JIRA_STATE_ON_HOLD}}` | Q5 — Jira workflow | Project state for "blocked / parked", e.g. `On Hold`, `Blocked` |
| `{{JIRA_TRANSITION_IN_PROGRESS}}` | Q9 — Jira transition IDs | Numeric ID for "In Progress" / "Work is started" in the project's workflow |
| `{{JIRA_TRANSITION_TESTING}}` | Q9 — Jira transition IDs | Numeric ID for "Testing" / "In Review" — used when PR is submitted |
| `{{JIRA_TRANSITION_DONE}}` | Q9 — Jira transition IDs | Numeric ID for "Done" / "Closed" |
| `{{JIRA_TRANSITION_ON_HOLD}}` | Q9 — Jira transition IDs | Numeric ID for "On Hold" / "Blocked" |
| `{{LABELS_SOURCE_OF_TRUTH_NOTE}}` | Developer confirms | One line pointing at the project's labels-matrix doc if one exists (e.g. a Confluence page URL), or "No shared labels matrix — apply your best judgment and confirm new labels with the team" if not |
| `{{LABEL_CATEGORY_1}}` / `_1_VALUES` | Developer confirms | E.g. `Stream` → `FE · BE · Data · DevOps · QA`. Add more category rows as needed; drop the placeholder rows entirely if the project has no categorized taxonomy beyond the harness-standard lifecycle labels |
| `{{LABEL_CATEGORY_2}}` / `_2_VALUES` | Developer confirms | E.g. `Version` → `V1 · V2 · MVP` |
| `{{DESCRIPTION_TEMPLATE_SOURCE_NOTE}}` | Developer confirms | One line pointing at the full description-template spec doc if one exists, else "Full spec: use the skeleton below directly." |
| `{{TEAM_COMPOSITION_NOTE}}` | Developer confirms | Either a pointer to a team-composition reference doc ("Source of truth: {{TEAM_COMPOSITION_PAGE_URL}} — consult when deciding assignment/ownership; never hardcode names here") or, if none exists, delete the whole "Team composition" section |
To find transition IDs: in Jira, open any issue in the project → Transitions API: `GET /rest/api/3/issue/{key}/transitions`.
## Conditionals
- **Jira Server / Data Center** (not Cloud): replace `accountId` assignment notes with `name`/`key` field — email-based lookup fails silently on Server.
- **No Confluence**: remove the Confluence URL line and the "return reference-doc URLs" note from the Fetch operation.
- **No labels matrix / simple taxonomy**: drop the `{{LABEL_CATEGORY_*}}` rows, keep only the harness-standard lifecycle row.
- **Different issue types**: if the project uses `Bug`, `Epic`, or custom types instead of `Task`/`Story`, update the Create operations accordingly.
- **No team-composition doc**: delete the "Team composition" section entirely.
## Verify checklist
- [ ] `{{PROJECT_KEY}}` resolved — no literal placeholder in issue key examples or error messages
- [ ] `{{JIRA_URL}}` and `{{CONFLUENCE_URL}}` resolved — valid URLs with no trailing slash
- [ ] All four transition IDs filled — verify against the project's actual Jira workflow (not assumed)
- [ ] All four project state names filled from the real workflow (`GET /rest/api/3/issue/{key}/transitions` returns both the `name` and the `id`); the generic column is left as shipped so dev/design requests by generic name still resolve
- [ ] `tools:` names the project's actual Jira MCP server in the `mcp__<server>__*` shape; an existing organisation connection is reused rather than a second `atlassian` server being assumed
- [ ] Transition IDs table matches the project's actual workflow states
- [ ] `tools:` frontmatter includes `mcp__atlassian__*` (or the deployment's actual Jira MCP tool namespace) — `Read` alone cannot call Jira
- [ ] Link direction verification section present verbatim — this is a real MCP tool behavior bug, not project-specific content, and should ship with every deployment using this Jira MCP integration
- [ ] Story description quality section forbids process/history narrative ("split from," "originally bundled with") — WHAT/WHY only
- [ ] `dev-plan/SKILL.md` delegates Jira start/fetch to `@scrum-master` (not inline MCP calls), if `{{provider:invoke:dev}}` in this deployment touches Jira in Plan
- [ ] `dev/SKILL.md` Completion delegates worklog + Done transition to `@scrum-master`
- [ ] `dev-document/SKILL.md` delegates TD creation to `@scrum-master`, if applicable
- [ ] `design/references/07-publish.md` tracker-sync step delegates worklog + Done transition to `@scrum-master` by state name only (no literal "Jira" wording, no transition IDs)
- [ ] `design/references/01-frame.md` Steps 2–3 delegate tracker fetch and the start-of-work transition to `@scrum-master` by state name only — this is a built-in default (not an optional deployment add-on); if Q5 is "none" (no issue tracker), these steps are removed entirely per `design/INTENT.md`'s conditionals
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 |