team-issues
Skill0 starsChanged 4 months agoMITView on GitHub
Import GitHub/Jira/Linear issues into the local backlog. Use with /team-issues.
In this repository42 other files it gives its agents.
- .claude/CLAUDE.mdCLAUDE.md
- content-rewriteSkill
- content-writeSkill
- core-auditSkill
- core-bootstrapSkill
- core-commitSkill
- core-explainSkill
- core-finalizeSkill
- core-orchestrateSkill
- core-pullSkill
- core-setupSkill
- core-updateSkill
- design-convertSkill
- design-shipSkill
- design-tokensSkill
- dev-inspectSkill
- dev-learnSkill
- dev-manualSkill
- dev-optimizeSkill
- dev-securitySkill
- dev-shipSkill
- dev-tweakSkill
- game-debugSkill
- game-optimizeSkill
- game-shipSkill
- game-tweakSkill
- marketing-contentSkill
- marketing-researchSkill
- marketing-screenshotsSkill
- project-addSkill
- project-appSkill
- project-memorySkill
- project-planSkill
- project-removeSkill
- project-researchSkill
- project-seedSkill
- project-syncSkill
- project-todoSkill
- project-tunnelSkill
- team-outsourceSkill
- team-reviewSkill
- team-verifySkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
15 rules. About 0 are common, 3 are distinctive, and 12 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
12
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 15
Import issues from a team tracker into your local backlog.
- Very specific: line 38
Team-mode gate: Read .project/project.json#team.mode (see shared/PROJECT-MODE.md).
- Very specific: line 54
Check team.tracker — if set, use that.
- Very specific: line 55
If not set: try gh repo view --json nameWithOwner (success → tracker = "github").
- Very specific: line 56
If that also fails, or if --paste flag is present → tracker = "paste".
- Very specific: line 59
Read .project/backlog.json → parse JSON → data.
- Very specific: line 61
If argument is a number (e.g.
- Very specific: line 108
If directIssueId is set → select that issue directly and skip the multi-select.
- Very specific: line 178
Single todo: ask via combined AskUserQuestion — Question 2 is two sequential modals (4-option cap per shared/SKILL-PATTERNS.md § Modal …
- Very specific: line 291
Write data.updated to today (YYYY-MM-DD).
- Very specific: line 293
Write project.json#team.tracker if it was not yet set.
- Distinctive: line 327
No bidirectional sync — issue updates in the tracker are not automatically reflected in the backlog
- Distinctive: line 330
Do NOT write code, do NOT run git commands
- Very specific: line 344
NEVER use blockquote syntax (>) for displaying content in user-facing output
- Very specific: line 349
Follow skills/shared/LANGUAGE.md for output language rules.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Team Issues
Import issues from a team tracker into your local backlog. Smart-splits multi-area issues into separate frontend/dev todos so each item maps cleanly to one skill pipeline.
Trigger: /team-issues, /team-issues <number>, /team-issues --mine, /team-issues --paste
Process
Seed task list at start:
TaskCreate with phases:
- PHASE 0: Pre-flight + tracker detection
- PHASE 1: Issue intake
- PHASE 2: Dedup + selection
- PHASE 3: Smart split analysis
- PHASE 4: User confirmation
- PHASE 5: Fragment confirm + checkpoint
- PHASE 6: Write to backlog
- PHASE 7: Output
PHASE 0: Pre-flight + tracker detection
Todo: mark PHASE 0 →
in_progress.
Team-mode gate. Read
.project/project.json#team.mode(seeshared/PROJECT-MODE.md). If"solo"or absent → show AskUserQuestion (warn-only):header: "Solo project" question: "This project is marked solo (team.mode). /team-issues is meant for projects with multiple contributors. Continue anyway?" options: - label: "Cancel (Recommended)" description: "Exit. Toggle to team via the ⚙ button in the backlog or run /core-setup to mark this as a team project." - label: "Yes, continue once" description: "Proceed with issue import for this single invocation." multiSelect: falseCancel → exit. Continue → proceed with PHASE 0 step 1.
- Read
.project/project.json:- Check
team.tracker— if set, use that. - If not set: try
gh repo view --json nameWithOwner(success →tracker = "github"). - If that also fails, or if
--pasteflag is present →tracker = "paste". - Save tracker to
project.json#team.trackerif it was missing.
- Check
- Read
.project/backlog.json→ parse JSON →data. - If argument is a number (e.g.
/team-issues 42) → store asdirectIssueId, skip PHASE 1 and 2.
Todo: mark PHASE 0 →
completed, PHASE 1 →in_progress.
PHASE 1: Issue intake
Todo: mark PHASE 1 →
in_progress.
GitHub:
gh issue list --state open --json number,title,body,labels,assignees,url,updatedAt --limit 50Extra filters:
--mineflag → add--assignee @me--label <label>arg → add--label <label>
Jira / Linear / paste:
AskUserQuestion:
header: "Issue intake"
question: "Paste the issue URL or description below."
options:
- label: "Paste URL"
description: "E.g. https://company.atlassian.net/browse/JIRA-456"
- label: "Paste text"
description: "Title + body — first line becomes the title"
multiSelect: falseThen ask for the content via free text. Parse:
- URL → extract
idvia regex (/([A-Z]+-\d+)/or/(\d+)$/),typefrom domain - Body → first line = title, rest = body
Todo: mark PHASE 1 →
completed, PHASE 2 →in_progress.
PHASE 2: Dedup + selection
Todo: mark PHASE 2 →
in_progress.
Filter already-imported issues — an issue is considered imported as soon as one backlog item has externalRef.id === issue.id && externalRef.type === tracker. Do not show those again in the selection.
If directIssueId is set → select that issue directly and skip the multi-select.
Otherwise: show AskUserQuestion (multiSelect=true) with per issue:
#42 Implement OAuth login [enhancement, P1] — @teammate
#38 Fix signup validation error [bug, P2]
#35 Add dashboard export [feature, P2]Todo: mark PHASE 2 →
completed, PHASE 3 →in_progress.
PHASE 3: Smart split analysis
Todo: mark PHASE 3 →
in_progress.
Per selected issue: analyze body inline for split signals.
Detection signals (in order of reliability):
- Headings (
## Frontend,## Backend,## Database,## API,## Tests,## Mobile) - Labeled lists (
- [ ] API endpoint,- [ ] UI component) - Keyword clusters in bullet points (
page,endpoint,migration,schema,component,test) - Length: body > 500 chars with ≥3 paragraph breaks → possibly splittable
Type mapping per fragment:
| Signal | Suggested type | Track |
| Frontend / UI / page / component | PAGE or COMPONENT | Design |
| Backend / API / endpoint / service | API or FEATURE | Dev |
| Database / migration / schema | FEATURE | Dev |
| Tests / test coverage | FEATURE | Dev |
| Bug fix | BUG | Dev |
| Unknown / mixed | FEATURE | Dev |
ASCII diagram: generate a visual split tree for the user so they can see at a glance what is proposed:
Issue #42: Implement OAuth login
├── oauth-login PAGE · Design · P1
├── oauth-callback API · Dev · P1
└── oauth-tests FEATURE · Dev · P2For a single fragment → skip the tree, go directly to PHASE 4 single-todo path.
Todo: mark PHASE 3 →
completed, PHASE 4 →in_progress.
PHASE 4: User confirmation per issue
Todo: mark PHASE 4 →
in_progress.
AskUserQuestion (single select):
header: "Import approach"
question: "How do you want to import this issue?"
options:
- label: "Smart split (Recommended)"
description: "Multiple todos based on the analysis"
- label: "Single todo"
description: "One backlog item for the entire issue"
- label: "Skip"
description: "Do not import this issue"
multiSelect: falseSmart split: go to PHASE 5.
Single todo: ask via combined AskUserQuestion — Question 2 is two sequential modals (4-option cap per shared/SKILL-PATTERNS.md § Modal Option Cap; show modal 2b only if "More →" is picked):
# Question 1
header: "Priority"
question: "Which priority?"
options:
- label: "P1 (Recommended)", description: "Highest priority"
- label: "P2", description: "Important, not blocking"
- label: "P3", description: "When there's time"
- label: "P4", description: "Park for later"
multiSelect: false
# Question 2a
header: "Type"
question: "Item type?"
options:
- label: "FEATURE (Recommended)"
- label: "BUG"
- label: "API"
- label: "More →"
multiSelect: false
# Question 2b (only if "More →" picked)
header: "Type"
question: "Item type?"
options:
- label: "PAGE"
- label: "CHANGE"
multiSelect: falseThen go directly to PHASE 6 (write).
Skip: go to next selected issue (loop back to PHASE 3 if more issues are waiting).
Todo: mark PHASE 4 →
completed, PHASE 5 →in_progress.
PHASE 5: Fragment confirm + checkpoint
Todo: mark PHASE 5 →
in_progress.
Show the split proposals as multi-select (default: all checked):
header: "Fragments"
question: "Which fragments do you want to import?"
options:
- label: "oauth-login · PAGE · Design · P1"
- label: "oauth-callback · API · Dev · P1"
- label: "oauth-tests · FEATURE · Dev · P2"
multiSelect: truePer selected fragment: show combined confirm for type + phase (pre-filled with proposal, user can override).
Interview checkpoint (before PHASE 6 writes):
Show summary table of everything being imported:
IMPORT OVERVIEW
════════════════════════════════════════════════════
Issue #42 · Implement OAuth login (github)
oauth-login PAGE · Design · P1
oauth-callback API · Dev · P1
oauth-tests FEATURE · Dev · P2
Issue #38 · Fix signup validation error (github)
signup-validation BUG · Dev · P2
════════════════════════════════════════════════════AskUserQuestion:
header: "Confirm import"
question: "Does this overview look correct? Items will be written to the backlog."
options:
- label: "Yes, import (Recommended)"
- label: "Adjust"
description: "Go back to selection"
multiSelect: falseTodo: mark PHASE 5 →
completed, PHASE 6 →in_progress.
PHASE 6: Write to backlog
Todo: mark PHASE 6 →
in_progress.
Per accepted todo: insert into data.features[] (after dedup check on name):
{
"name": "{kebab-case of title or fragment label}",
"type": "{type}",
"status": "TODO",
"phase": "{phase}",
"description": "{issue title}\n\n{fragment excerpt, max 500 chars}",
"source": "/team-issues",
"origin": "user",
"externalRef": {
"type": "{github|jira|linear}",
"id": "{issue id}",
"url": "{url or null for paste without URL}",
"labels": ["{label.name}"],
"split": "{frontend|backend|tests|null}"
},
"dependencies": []
}Write data.updated to today (YYYY-MM-DD).
Write project.json#team.tracker if it was not yet set.
Edit the JSON in .project/backlog.json (see shared/BACKLOG.md § Writing).
Todo: mark PHASE 6 →
completed, PHASE 7 →in_progress.
PHASE 7: Output
Todo: mark PHASE 7 →
in_progress.
IMPORTED
Issue #42: Implement OAuth login (github)
oauth-login P1 · PAGE · Design track
oauth-callback P1 · API · Dev track
oauth-tests P2 · FEATURE · Dev track
Issue #38: Fix signup validation error (github)
signup-validation P2 · BUG · Dev track
Backlog: .project/backlog.json
Next steps:
- /dev-ship oauth-callback (start dev pipeline)
- /dev-ship signup-validation (start dev pipeline)
- /design-convert oauth-login (start design pipeline)Todo: mark PHASE 7 →
completed.
Restrictions
- Smart split is a suggestion — the user decides which fragments are ultimately imported
- No bidirectional sync — issue updates in the tracker are not automatically reflected in the backlog
- Smart-split one issue at a time — no bulk-split of multiple issues at once
- Dedup on
externalRef.id+ tracker — an issue that is already (partially) imported does not appear again in the selection - Do NOT write code, do NOT run git commands
Tracker-support matrix
| Tracker | Method | ID format |
| GitHub | gh CLI | #123 |
| Jira | paste-flow | PROJ-456 |
| Linear | paste-flow | ABC-789 |
Native Jira/Linear integration (API/MCP) is out of scope for v1. See shared/TEAM.md for Jira/Linear manual workflow.
Terminal Formatting
- NEVER use blockquote syntax (
>) for displaying content in user-facing output - Backticks only for actual code, file paths, and command references
Language
Follow skills/shared/LANGUAGE.md for output language rules.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

