Skip to content
cccskills
FreePublished on GitHub

daily-briefing-builder

Generate a clean morning brief in Claude Code — pulls today's priorities, unposted content, and weather from your vault.

The English summary is being prepared. Showing the original description.

Run score Not tested

See how to install

Files included (2)

  • SKILL.md5.2 KB
  • SKILL-OC.md1.6 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Daily Briefing Builder

Generates a morning brief from your Obsidian vault. Reads today's action file, scans for unposted content, and fetches weather — all inside a Claude Code session.

No APIs. No paid services. No agent autonomy required. You run it, you get your brief.


How to Use

Open Claude Code in your vault directory and say:

Run the Daily Briefing Builder skill.

Or with context upfront:

Run the Daily Briefing Builder skill. My vault is at /path/to/vault and my city is Ann+Arbor.

Skill Instructions (for Claude Code)

When this skill is invoked, follow these phases exactly.


PHASE 1: INTAKE

Check whether the user has provided:

  • vault_path — absolute path to their Obsidian vault
  • city — city name for weather (wttr.in format, spaces as +)

If either is missing, ask before proceeding:

To run your morning brief, I need two things:
1. Your vault path (e.g. /root/obsidian-vault)
2. Your city for weather (e.g. Ann+Arbor or London)

Do not proceed to PHASE 2 until both values are confirmed.


PHASE 2: ANALYZE

Run these shell commands in sequence. Capture all output before formatting.

Step 1 — Today's actions:

TODAY=$(date +%Y-%m-%d)
VAULT="VAULT_PATH_HERE"
ACTIONS_FILE="$VAULT/bambf/tracking/daily-actions/${TODAY}.md"

if [ -f "$ACTIONS_FILE" ]; then
  echo "FILE_FOUND"
  awk '/## Today.s 3 Actions/{found=1; next} found && /^[0-9]/{print} found && /^##/{exit}' "$ACTIONS_FILE"
else
  echo "FILE_MISSING:$ACTIONS_FILE"
fi

Step 2 — Unposted content scan:

VAULT="VAULT_PATH_HERE"
READY_DIR="$VAULT/content/ready-to-post"

if [ -d "$READY_DIR" ]; then
  find "$READY_DIR" -name "*.md" -printf "%T@ %p\n" 2>/dev/null \
    | sort -rn | awk '{print $2}' \
    | while read f; do
        if grep -q '\*\*Posted:\*\* ❌' "$f" 2>/dev/null; then
          platform=$(echo "$f" | sed "s|.*/ready-to-post/||" | cut -d'/' -f1)
          title=$(grep -m1 '^[^#\-\*>|` ]' "$f" 2>/dev/null | head -c 100)
          rel="${f##$VAULT/}"
          echo "ITEM|$platform|$rel|$title"
        fi
      done
else
  echo "DIR_MISSING:$READY_DIR"
fi

Step 3 — Weather:

curl -s --max-time 5 "wttr.in/CITY_HERE?format=3" || echo "WEATHER_UNAVAILABLE"

PHASE 3: OUTPUT

Format and print the brief to the terminal:

☀️ Morning Brief — [Weekday, Month Day]

TODAY'S 3 ACTIONS
[numbered actions, or fallback message if file missing]

READY TO POST ([shown] of [total])
[platform] [title] — [relative file path]
...and X more in the pipeline   ← only if total > 5

WEATHER
[wttr.in output or fallback]

Formatting rules:

  • Cap ready-to-post display at 5 items (newest-modified first)
  • If content/ready-to-post/ doesn't exist: No ready-to-post folder found
  • If no unposted files: Content queue is empty
  • If actions file missing: No actions file for today — create one at bambf/tracking/daily-actions/YYYY-MM-DD.md
  • If weather fails: Weather unavailable (offline) — do not abort the brief

PHASE 4: SELF-CRITIQUE

Before delivering, run this internal check. Fix anything that fails — do not deliver until all 4 pass.

  1. Actions accuracy — Was the actions file found? Did numbered items extract correctly? If the file exists but no items were found, note the potential heading mismatch.
  2. Content completeness — Count the ITEM| lines in Step 2 output. Does the number shown in the brief match? Were any files skipped?
  3. Weather result — Did curl return valid output? Is the fallback shown if it failed?
  4. Format cleanliness — No raw bash output. No placeholder text. All three sections present.

Example Output

☀️ Morning Brief — Friday, February 28

TODAY'S 3 ACTIONS
1. Finish Q1 content calendar
2. Send invoice to BAMBF client
3. Publish LinkedIn post on AI ops

READY TO POST (5 of 9)
[linkedin] Nobody talks about what happens when your AI system fails... — content/ready-to-post/linkedin/ai-ops-failure.md
[twitter] The overnight ops shift is real. We run Scribe at 2am... — content/ready-to-post/twitter/async-ops.md
[newsletter] AI marketing systems that actually work — content/ready-to-post/newsletter/systems-post.md
[linkedin] Three things I stopped doing after deploying agents — content/ready-to-post/linkedin/stopped-doing.md
[twitter] Founders who ignore async AI ops are about to find out why — content/ready-to-post/twitter/ignore-ops.md
...and 4 more in the pipeline

WEATHER
Ann Arbor: ☀️  +42°F

Daily Actions File Format

Expected path: <vault>/bambf/tracking/daily-actions/YYYY-MM-DD.md

Expected structure:

# Daily Actions — 2026-03-01

## Today's 3 Actions
1. First priority
2. Second priority
3. Third priority

If you use a different path or heading, tell Claude Code when invoking the skill.


Requirements

  • Claude Code with bash tool access
  • Vault with content/ready-to-post/ folder (files must contain **Posted:** ❌ to register as unposted)
  • curl installed (weather fails gracefully without it)
  • Daily actions file optional — brief still runs without it

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Audit how a brand appears in AI-powered search (ChatGPT, Perplexity, Claude, Gemini). Use when user mentions "AI search," "how do I show up in ChatGPT," "AI discoverability," "AEO," "LLM visibility," or wants to understand their brand's AI presence.

The English summary is being prepared. Showing the original description.

BrianRWagner/ai-marketing-claude-code-skills444Updated Mar 19, 2026

Turn client wins into formatted case studies for proposals, social proof, and sales conversations. Use when someone needs to document results, build credibility, or create reusable proof assets.

The English summary is being prepared. Showing the original description.

BrianRWagner/ai-marketing-claude-code-skills444Updated Mar 19, 2026

Build personalized cold outreach sequences for LinkedIn and email. Use when someone needs to reach prospects, warm up cold leads, or build a systematic outreach engine. Covers research, connection requests, follow-ups, and conversion.

The English summary is being prepared. Showing the original description.

BrianRWagner/ai-marketing-claude-code-skills444Updated Mar 19, 2026

Generate content ideas rooted in positioning. Use when someone needs "content ideas," "what should I post," "blog topics," "LinkedIn ideas," or is stuck on what to create.

The English summary is being prepared. Showing the original description.

BrianRWagner/ai-marketing-claude-code-skills444Updated Mar 19, 2026

Remove AI-generated jargon and restore human voice to text. Built from analyzing 1,000+ AI vs human content pieces.

The English summary is being prepared. Showing the original description.

BrianRWagner/ai-marketing-claude-code-skills444Updated Mar 19, 2026

Autonomous goal execution — give a goal, get a plan, confirm, execute, report. You steer, Claude drives.

The English summary is being prepared. Showing the original description.

BrianRWagner/ai-marketing-claude-code-skills444Updated Mar 19, 2026

See all skills by BrianRWagner

Report a problem with this skill