morning-triage

name	morning-triage
description	Scan DMs, @mentions, Slack channels, and Jira to create a prioritized daily plan. Run this every morning to start the day focused.

Path resolution: This skill may run from any repo. All context/ and config.yaml paths are relative to the workbench root, not the current working directory. Read ~/.codex/workbench-root or ~/.claude/workbench-root to get the absolute workbench path, then prepend it to all context/ and config.yaml references. See PATHS.md.

Morning Triage

Scan all relevant channels, check DMs and @mentions, check Jira, and create a prioritized daily plan.

The user's display name is defined in config.yaml under user.slack_display_name. Slack channels are defined in config.yaml under slack.channels. Jira projects are defined in config.yaml under jira.projects.

Process

Step 1: Check DMs and @Mentions

This is the highest priority step. Before anything else:

1. Read config.yaml to get the user's Slack display name from user.slack_display_name.
2. Search for recent messages mentioning that display name using mcp__claude_ai_Slack_MCP__slack_search_public_and_private.
3. Also search with query: to:me or check DMs for unread messages.

For each DM or @mention found:

• Read the full thread using mcp__claude_ai_Slack_MCP__slack_read_thread
• Classify as: needs response, FYI, or already handled
• Note who sent it and how urgent it appears

Step 2: Read Slack Channels

Read channels from config.yaml: slack.channels for messages from the last 24 hours. Focus on threads that are unresolved, need input, or signal something urgent.

Check channels in category order (operations first for incidents, then product, then engineering, then general). Use mcp__claude_ai_Slack_MCP__slack_read_channel for each channel.

Step 3: Check Jira

Search for tickets that need attention using the project keys from config.yaml: jira.projects:

1. Tickets assigned to me that are in progress or to-do (use searchJiraIssuesUsingJql with all configured project keys)
2. Tickets updated in the last 24 hours in all configured projects
3. Any tickets in "Blocked" status

Step 4: Check Yesterday's Plan and Weekly Plan

• Read context/active/daily.md for yesterday's plan (if the date in it is a previous day, it hasn't been archived yet). Identify anything that wasn't completed and should carry forward.
• Read context/active/weekly.md for this week's capacity state. Note remaining capacity and any items that are due this week.
• If no weekly plan exists, flag this and suggest creating one after the daily triage.

Step 5: Get Today's Calendar

Check context/active/calendar.md for today's actual meeting schedule.

• If the file exists, use the real meeting hours and available hours for today.
• If the file is missing or outdated, suggest running /calendar-sync first.
• Note the best focus blocks (longest meeting-free stretches) for scheduling deep work.

Step 6: Estimate and Prioritize

For each task identified (DMs, Jira tickets, carried items), create a quick time estimate:

• Read context/calibration.md for calibrated multipliers and pace factor
• Apply calibrated multipliers if available, otherwise use defaults from config.yaml: estimation.multipliers
• Keep estimates lightweight at this stage — gut feel with the calibrated minimum multiplier
• Sum up the total estimated hours

Check against today's real capacity (from calendar):

Meetings today: Xh (from calendar-sync)
Available today: Yh (6h minus meetings, adjusted for pace factor)
Total estimated: Zh
Status: [fits / tight / overcommitted]
Best focus block: HH:MM - HH:MM (Xh uninterrupted)

If overcommitted, explicitly flag which items should be deferred or delegated.

Step 7: Archive Yesterday and Create Today's Plan

Before creating today's plan:

1. Check if context/active/daily.md exists and its date header is from a previous day.
2. If so, move it to context/archive/YYYY/MM/daily/YYYY-MM-DD.md (creating directories as needed).
3. Then create a fresh context/active/daily.md.

Structure the plan with:

1. Capacity Today — available hours after meetings, committed hours, remaining.
2. DMs and @Mentions — anything that needs a direct response, with time estimates. Listed first since these are people waiting on you specifically.
3. Top Priorities (max 5) — ordered by impact. For each item: WHY it matters, WHO is affected, and estimated hours.
4. Carried Forward — incomplete items from yesterday, if still relevant, with estimates.
5. Slack Threads Needing Response — specific threads where input is needed.
6. Meetings Today — if mentioned in Slack or known.

If the total estimated hours exceed available capacity, call this out explicitly and recommend what to defer.

Keep context/active/daily.md as the automation source of truth. If the daily plan is meant to be reviewed or shared beyond the session, also create/update context/active/daily.html using context/standards/html-plan-standard.md as a companion view. Do not replace daily.md.

Step 8: Present the Summary

Give a brief verbal summary:

• X DMs / @mentions waiting for your response
• Y items from channels need attention
• Z Jira tickets are active
• W items carried from yesterday
• Total estimated: Xh against Yh available
• Capacity verdict: [fits / tight / overcommitted — need to cut Z]
• Here's the recommended priority order and why

Ask if the priorities look right or if anything should be reordered or deferred.

Important Notes

• DMs and @mentions come first. Someone directly reaching out to you is always higher signal than channel noise.
• Be concise. This should take under 2 minutes to read.
• Lead with action items, not status reports.
• If something is genuinely urgent (incident, blocked deployment, customer escalation), flag it at the very top before the normal plan.
• Don't include noise — skip channels with nothing actionable.