agentsclimarketplace

Meeting summarizer

Skill JayRHa/AgentSkills/meeting-summarizer

The largest community-driven library of Agent Skills (SKILL.md + scripts/references/examples) for Claude, Codex, Gemini CLI, Cursor and friends.

Install
npx -y skills add JayRHa/AgentSkills --skill meeting-summarizer

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Transforms raw meeting notes, transcripts, or recordings text into a concise, scannable summary with clear sections for decisions made and owned action items (owner + due date). Use this skill when the user asks to "summarize this meeting", "turn these notes into a summary", "extract action items", "who agreed to what", "write meeting minutes", "TLDR this transcript", "pull decisions and next steps", or pastes a Zoom/Teams/Google Meet transcript and wants it condensed. Also applies to standups, 1:1s, planning sessions, retros, and client calls.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

6.7 KB, as published. Nobody here has run it

Meeting Summarizer

Overview

This skill converts messy meeting input (chat transcripts, dictated notes, bullet dumps, recording transcripts) into a structured, decision-oriented summary that a busy reader can absorb in under a minute. The output prioritizes the two things people actually need after a meeting: what was decided and who owes what by when.

Keywords: meeting minutes, summary, recap, action items, decisions, owners, due dates, TLDR, standup, retro, 1:1, transcript, notes.

The non-negotiable rule: every action item must have an owner. If the source does not name one, the action item is flagged as OWNER: UNASSIGNED rather than silently dropped or guessed.

When To Use

Use this skill whenever the user wants raw meeting content distilled. Triggers include pasting a transcript, asking for "minutes", "a recap", "key takeaways", "next steps", or "the TLDR". Do not use it for live note-taking guidance or scheduling — only for processing content that already exists.

Process

Follow these steps in order.

  1. Detect input type & length. Identify whether the input is a verbatim transcript (timestamps, speaker labels), shorthand notes, or a mixed dump. Note the approximate length so you can size the summary.

  2. Identify metadata. Pull (or infer where stated) the meeting title/topic, date, and attendees. Never invent a date — if absent, omit it or write Date: not stated.

  3. Extract the four content streams by scanning the whole input once for each:

    • Decisions — anything settled, approved, chosen, or rejected. Look for trigger words: "we'll go with", "agreed", "approved", "decided", "let's do", "no to", "final".
    • Action items — future work someone committed to. Look for: "I'll", "can you", "take", "follow up", "by Friday", "next step", assigned verbs.
    • Discussion / context — the reasoning, options weighed, and notable points that aren't decisions or actions.
    • Open questions / risks — unresolved items, blockers, parking-lot topics.
  4. Resolve owners and dates for each action item. Map pronouns to names using the transcript ("I'll handle it" said by Maria → owner: Maria). Convert relative dates ("by next Tuesday") to absolute dates only if a meeting date is known; otherwise keep the relative phrasing. Mark missing owners UNASSIGNED.

  5. Deduplicate and compress. Merge restated points. Cut filler, small talk, and off-topic tangents. Each bullet should carry one idea, written as a terse, complete statement.

  6. Assemble the summary using the structure in templates/summary.md. Lead with a 1–3 sentence TLDR, then Decisions, then Action Items (as a table), then Discussion, then Open Questions. Omit any empty section except Action Items (which always renders, even if to state "None captured").

  7. Quality-check against references/checklist.md before delivering. Most importantly verify the no-orphan-actions rule and that no decision was downgraded to a discussion bullet.

Output Structure

# <Meeting Topic> — Summary
Date: ... | Attendees: ...

## TLDR
1–3 sentences.

## Decisions
- ...

## Action Items
| # | Action | Owner | Due |

## Discussion Highlights
- ...

## Open Questions / Risks
- ...

See templates/summary.md for the full fill-in template and examples/standup-recap.md for a complete worked example.

Decision Framework: classifying a line

Use this quick test on each notable line:

  • Is it a settled choice or approval? → Decision.
  • Does it commit a specific person to future work? → Action Item.
  • Is it unresolved, a blocker, or a "we should look into…"? → Open Question / Risk.
  • Is it reasoning, an option considered, or notable context? → Discussion Highlight.
  • Is it filler, greeting, or off-topic? → Drop.

When something is both a decision and spawns work (e.g., "We'll launch Tuesday — Sam, prep the release notes"), record the decision in Decisions AND the task in Action Items.

Adapting Length & Tone

  • Short input (<1 page): keep the whole summary to ~10 bullets total.
  • Long transcript (>30 min): group Discussion Highlights under sub-topics; keep TLDR to 3 sentences max.
  • Standup: collapse to per-person blockers + the day's commitments.
  • Client/external call: neutral, professional tone; spell out acronyms once; avoid internal jargon.
  • Retro: organize Discussion under Went Well / Didn't / Try.

For tone, owner-resolution edge cases, and a helper script, see the references below.

Helper Script

scripts/extract_actions.py is a zero-dependency Python tool that scans a notes/transcript file for likely action-item and decision lines using cue-word heuristics, and prints them grouped — a fast first pass before you write the polished summary. Run:

python3 scripts/extract_actions.py meeting.txt
python3 scripts/extract_actions.py meeting.txt --json

It is a draft aid, not a replacement for reading the full input.

Best Practices

  • Lead with the TLDR; many readers stop there.
  • Write action items as imperatives starting with a verb ("Send the revised deck", not "deck needs sending").
  • Always pair owner + due date; if either is missing, mark it explicitly.
  • Preserve exact figures, dates, names, and dollar amounts verbatim — never round or paraphrase numbers.
  • Use the attendees list to disambiguate first-name-only references.
  • Keep the reader's perspective: what do they need to do or know.

Common Pitfalls

  • Orphan actions: dropping a task because no owner was named. Flag UNASSIGNED instead.
  • Hallucinated owners/dates: never assign a person or deadline the source didn't state.
  • Burying decisions in prose: decisions belong in their own section, not mixed into discussion.
  • Over-summarizing: losing a key number, name, or conditional ("only if budget approved").
  • Copying filler: greetings, "can you hear me?", scheduling chatter — drop it.
  • Inventing a date: if the meeting date isn't in the source, don't fabricate one or convert relative dates against today.

See references/checklist.md for the full pre-delivery QA list and references/owner-resolution.md for handling tricky attribution cases.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.