agentleFS
jcesarperezjcesarperez/claude-em/.claude/skills/check-board/SKILL.md

check-board

Skill94 starsChanged 5 months agoMITView on GitHub

Evaluate the daily health of a Jira Kanban board: flow, people load, and bugs at risk. Use whenever the user wants a daily-standup-ready view of a board — what's stuck, who's overloaded, who can pull, what's blocked, what bugs are about to breach SLA.

The file

Show its rules and text

We split this skill into the rules it gives an agent

20 rules. About 0 are common, 9 are distinctive, and 11 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 11

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run this script using the Bash tool:

    line 11
  2. Very specific:

    If the team file is missing or has no Default board, ask the user for the board id before continuing.

    line 67
  3. Very specific:

    Run the script with --json.

    line 72
  4. Very specific:

    Use the pre-computed metrics for all rubric checks.

    line 73
  5. Very specific:

    If flagged_field_supported == false, mention in the output that blocked detection is unavailable for this tenant.

    line 75
  6. Very specific:

    If flagged_field_supported == false: skip this signal and say so explicitly.

    line 99
  7. Distinctive:

    Do NOT flag unassigned issues in To Do — that's normal in this team.

    line 122
  8. Very specific:

    Filter to the team roster: only include names that match team_roster (full names from the team file).

    line 126
  9. Distinctive:

    If no team file was loaded (board id passed directly), skip this signal and write "Idle filter unavailable — no team file resolved."

    line 127
  10. Very specific:

    Use metrics.epics. Each epic groups all active (non-Done) issues whose parent.issuetype == "Epic".

    line 137
  11. Very specific:

    If issues_without_epic > 0, mention the count at the end so the EM knows that group exists.

    line 146
  12. Distinctive:

    All epics OK | (only when no epic is 🔴/🟡)

    line 203
  13. Distinctive:

    Always cite issue keys, durations, dates, and people by name — no vague observations.

    line 234
  14. Distinctive:

    DO NOT flag unassigned issues in To Do — only in In Progress.

    line 242
  15. Very specific:

    DO NOT list as idle anyone outside team_roster when a team file was loaded (PMs, designers, cross-team contributors stay off the line).

    line 243
  16. Very specific:

    DO NOT recompute metrics from issues.

    line 244
  17. Distinctive:

    DO NOT show Done items older than 36h (the script already trims them).

    line 245
  18. Distinctive:

    DO NOT use sprint/velocity language — this skill is Kanban-only.

    line 246
  19. Very specific:

    If flagged_field_supported == false, say "Blocked detection unavailable" once and skip the Blocked block entirely.

    line 247
  20. Distinctive:

    Avoid generic recommendations like "improve communication".

    line 248

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Tool

Run this script using the Bash tool:

bash .claude/skills/check-board/scripts/fetch_board.sh <BOARD_ID_OR_URL> --json
  • BOARD_ID_OR_URL: numeric board id (e.g. 5289) or full board URL.
  • The script trims Done to the last 36h. To Do, In Progress and Done<36h all show up in the snapshot.

The script outputs JSON with this structure:

{
  "fetched_at": "2026-05-09T12:00:00Z",
  "board": {
    "id": "5289",
    "name": "...",
    "type": "kanban",
    "project": "OPS",
    "columns": [{ "name": "In Progress", "status_ids": ["10001", "10002"] }]
  },
  "flagged_field_supported": true,
  "metrics": {
    "total": 0,
    "in_progress": 0,
    "todo": 0,
    "done_last_36h": 0,
    "blocked_count": 0,
    "blocked": [{ "key": "...", "summary": "...", "status": "...", "statusCategory": "...", "column": "...", "assignee": "Name", "time_in_status": "Xd Yh Zm" }],
    "stuck": [{ "key": "...", "summary": "...", "status": "...", "column": "...", "assignee": "Name", "is_blocked": false, "time_in_status": "Xd Yh Zm", "seconds": 0, "severity": "warning|critical" }],
    "unassigned_in_progress": [{ "key": "...", "summary": "...", "status": "...", "column": "...", "time_in_status": "..." }],
    "regressions_last_7d": [{ "key": "...", "summary": "...", "at": "...", "from": "...", "to": "..." }],
    "throughput_7d": 0,
    "throughput_14d": 0,
    "median_cycle_time_days": 0.0,
    "columns": [{ "name": "...", "count": 0, "in_progress_count": 0, "stuck_count": 0 }],
    "assignees": [{ "assignee": "Name or null", "count": 0, "keys": ["..."], "max_time_in_status_seconds": 0, "stuck_count": 0, "blocked_count": 0 }],
    "idle_assignees": ["Name"],
    "bugs_at_risk": [{ "key": "...", "summary": "...", "status": "...", "column": "...", "assignee": "Name", "priority": "High", "duedate": "YYYY-MM-DD or null", "days_until_due": 0, "risk": "red|yellow", "reason": "due date|Highest priority, no due date" }],
    "bugs_other_count": 0,
    "epics": [{ "key": "EPIC-1", "summary": "...", "status": "...", "active_count": 0, "in_progress_count": 0, "stuck_count": 0, "blocked_count": 0, "red_bug_count": 0, "yellow_bug_count": 0, "issue_keys": ["..."] }],
    "issues_without_epic": 0
  },
  "issues": [/* full snapshot, same per-issue shape as inside metrics + changelog-derived fields */]
}

Instructions

Resolving the input

The skill accepts either:

  • A team name (e.g. events, floor) → read data/team_{name}.md, take the Default board and the team roster (full names of all ### member sections). Use that board id; keep the roster for filtering.
  • A board id or URL → use it directly. No roster available; skip roster-based filters and say so.

If the team file is missing or has no Default board, ask the user for the board id before continuing.

Analysis steps

  1. Resolve the input as above. If a team file was loaded, build team_roster = [full names from every "### name" section].
  2. Run the script with --json. Stop and report the error if it fails.
  3. Use the pre-computed metrics for all rubric checks. Do not recompute from issues unless you need a specific issue's detail (e.g. parent epic, blocked_by links).
  4. Produce the structured report below.
  5. If flagged_field_supported == false, mention in the output that blocked detection is unavailable for this tenant.

Scoring Rubrics

Flow Health

Evaluate four signals:

Stuck (use metrics.stuck):

  • 🟢 None
  • 🟡 1–2 warnings (≥5d), no critical
  • 🔴 Any critical (≥10d), or ≥3 warnings

Regressions last 7d (use metrics.regressions_last_7d):

  • 🟢 0
  • 🟡 1–2
  • 🔴 ≥3, or ≥1 transition into a To Do category from In Progress/Done

Blocked (use metrics.blocked_count and metrics.blocked):

  • 🟢 0
  • 🟡 1–2
  • 🔴 ≥3, or any blocked item also stuck ≥10d
  • If flagged_field_supported == false: skip this signal and say so explicitly.

Throughput / Cycle time:

  • Cite throughput_7d, throughput_14d, and median_cycle_time_days.
  • These are descriptive — flag only as 🟡/🔴 if you spot an obvious anomaly (e.g. throughput_7d == 0 while in_progress > 5, or median_cycle_time_days > 15). Otherwise 🟢.

Overall Flow state:

  • 🟢 Moving — all four 🟢
  • 🟡 At Risk — any 🟡, no 🔴
  • 🔴 Stalled — any 🔴

People Health

Evaluate three signals:

WIP balance (use metrics.assignees):

  • Overloaded: assignee with count ≥ 4 in In Progress, or count ≥ 3 AND stuck_count ≥ 3
  • Heavy stuck: assignee with stuck_count ≥ 3 regardless of count
  • Cite each by name.

Unassigned in In Progress (use metrics.unassigned_in_progress):

  • 🟢 None
  • 🔴 Any — active work without an owner is a delivery risk.
  • Do NOT flag unassigned issues in To Do — that's normal in this team.

Idle / can-pull (use metrics.idle_assignees):

  • Assignees with items elsewhere on the board but 0 in In Progress. Useful daily signal: who can pick something up.
  • Filter to the team roster: only include names that match team_roster (full names from the team file). This excludes PMs, designers, cross-team contributors and anyone else who shows up on the board but isn't an engineer on the team.
  • If no team file was loaded (board id passed directly), skip this signal and write "Idle filter unavailable — no team file resolved."
  • Note: people not assigned to anything on the board are invisible to this signal.

Overall People state:

  • 🟢 Healthy — no overload, no unassigned in progress
  • 🟡 At Risk — 1 overloaded person, or any single 🟡 signal
  • 🔴 Stretched — multiple overloaded, or any unassigned in progress, or someone with stuck_count ≥ 4

Epic Health

Use metrics.epics. Each epic groups all active (non-Done) issues whose parent.issuetype == "Epic". Sub-tasks whose parent is a Story are grouped under their Story key (treated as "(no epic)" in the listing) — their epic is unknown without an extra fetch.

For each epic, classify state:

  • 🔴 Stalled — stuck_count ≥ 3 OR blocked_count ≥ 2 OR red_bug_count ≥ 1
  • 🟡 At Risk — stuck_count ≥ 1 OR blocked_count ≥ 1 OR yellow_bug_count ≥ 1 OR in_progress_count ≥ 5 (high WIP)
  • 🟢 OK — none of the above

In the output, list only 🔴 and 🟡 epics (those that need attention). If all epics are 🟢, say "All epics OK". Always cite the epic's key + summary and the specific signals (e.g. "3 stuck, 1 blocked, 1 red bug").

If issues_without_epic > 0, mention the count at the end so the EM knows that group exists.

Bugs at Risk

Already classified by the script. Output them as-is, sorted (the script already orders red before yellow, then by days_until_due ascending).

  • 🔴 Red bugs: SLA imminent (days_until_due ≤ 3, including overdue) OR priority == Highest with no due date.
  • 🟡 Yellow bugs: due in 4–10d.
  • The bugs_other_count is informational ("Bugs without SLA: N") — do not flag.

Overall Bugs state:

  • 🟢 No red, no yellow
  • 🟡 Any yellow, no red
  • 🔴 Any red

Output Format

Return ONLY this structure:


📋 Board Snapshot — BOARD_ID: Name [type]

  • Total (snapshot): X · In Progress: X · To Do: X · Done last 36h: X
  • Blocked (Flagged): X | Flagged field not available on this tenant
  • Throughput: X (7d) · X (14d)
  • Median cycle time (last 14d): X.Xd | Not enough data
  • Columns: name (count), name (count), ...

🌊 Flow Health — 🟢 Moving / 🟡 At Risk / 🔴 Stalled

  • Stuck: 🟢/🟡/🔴 — X warnings, X critical
    • KEY — summary (Xd in "Status", assignee)[ · BLOCKED]
    • ...up to ~6, prioritize critical
  • Regressions last 7d: 🟢/🟡/🔴 — X transitions
    • KEY — summary: from → to (date)
    • ...
  • Blocked: 🟢/🟡/🔴 — X items
    • KEY — summary (status, assignee, Xd in status)
    • ...
  • Throughput: X / 7d, X / 14d — 🟢/🟡 (note anomaly if any)
  • Cycle time: X.Xd median — 🟢/🟡/🔴

👥 People Health — 🟢 Healthy / 🟡 At Risk / 🔴 Stretched

  • Overloaded: Name (X WIP, X stuck), ... | None
  • Heavy stuck: Name (X stuck), ... | None
  • Unassigned in In Progress: KEY — summary, KEY — summary | None
  • Can pull (idle): Name, Name | None visible

🛤️ Epic Health

  • 🔴 KEY — epic summary X active · X WIP · X stuck · X blocked · X 🔴 / X 🟡 bugs
  • 🟡 KEY — ...
  • All epics OK | (only when no epic is 🔴/🟡)
  • Issues not under an epic on the board: X (informational) | (skip if 0)

🐞 Bugs at Risk — 🟢 / 🟡 / 🔴

  • 🔴 KEY — summary priority · due YYYY-MM-DD (Xd) · assignee — reason
  • 🟡 KEY — summary priority · ...
  • ...
  • Bugs without SLA on the board: X (informational)

🧭 Summary

  • 3–5 bullets max. The actionable story for today's standup. Focus on:
    • The 1–2 issues most likely to slip today
    • Who needs help (overloaded, heavy stuck)
    • Who can pull (idle)
    • Top blocker to unblock
    • Any SLA bug due today/overdue

🎯 Recommendations

  1. [Verb] Specific concrete action — cite KEY or Name
  2. ... (2–5 items, ordered by urgency)

Style Guidelines

  • Always cite issue keys, durations, dates, and people by name — no vague observations.
  • Mention assignee on every stuck/blocked/at-risk item so the EM knows who to talk to.
  • For overdue bugs, lead with the overdue magnitude ("Xd overdue").
  • Be concise and direct — EM tone, no fluff.
  • Prioritize signal over completeness: if there are 20 stuck items, list the top ~6 most severe and say "and N more".

Constraints

  • DO NOT flag unassigned issues in To Do — only in In Progress.
  • DO NOT list as idle anyone outside team_roster when a team file was loaded (PMs, designers, cross-team contributors stay off the line).
  • DO NOT recompute metrics from issues; trust the pre-computed metrics.
  • DO NOT show Done items older than 36h (the script already trims them).
  • DO NOT use sprint/velocity language — this skill is Kanban-only.
  • If flagged_field_supported == false, say "Blocked detection unavailable" once and skip the Blocked block entirely.
  • Avoid generic recommendations like "improve communication". Be specific: "ping <name> on KEY-X — stuck Xd in Review with no assignee".

Goal

Help an Engineering Manager walk into the daily knowing:

"What's slipping, who's overloaded, who can pull, what's blocked, and which bug is about to breach SLA?"

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.