More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Meeting Action Extractor

    1

    Convert messy meeting transcripts, call notes, and chat threads into structured executive summaries, owner-assigned action items, and risk logs.

    Free

    2 uses5.0 (1 review)
    meeting-action-extractor

    meeting-action-extractor

    Example session with this skill installed

    Please process these notes from today's Sprint Planning (2023-10-25). Sarah: "We need the API docs done by Friday." Dave: "I'll take that." Sarah: "Should we use GraphQL?" Dave: "No, let's stick to REST for now." We also decided to delay the UI overhaul. Mark mentioned he's blocked by the database migration.

    • Read your context and instructions
    • Compiled the meeting-action-extractor

    Extraction complete. I have classified the API documentation as a P0 task and flagged the database migration as a critical blocker.

    Section 1 — Meeting Metadata & Signal Snapshot

    Meeting TopicDate/TimeParticipantsDecisionsAction Items (A/U)Blockers
    Sprint Planning2023-10-25Sarah, Dave, Mark21 (1/0)1

    Section 2 — 3-Sentence Executive TL;DR

    The team met to define priorities for the upcoming sprint and resolve architectural questions. We finalized the decision to use REST over GraphQL and postponed the UI overhaul to prioritize core backend stability. The immediate critical path requires Dave to complete API documentation by Friday to unblock downstream tasks. [48 words]

    Section 3 — Finalized Decisions Locked In

    Decision: Use REST API instead of GraphQL.

    Rationale: Efficiency and consistency with current stack. [Dave/Sarah, 2023-10-25]

    • Decision: Delay UI overhaul. Rationale: Focus resources on backend and documentation. [Group, 2023-10-25]

    Section 4 — Action Items Tracker

    IDPriorityAction ItemOwnerDue DateSource
    A1P0Complete API documentationDave2023-10-27[Sarah/Dave]

    Section 5 — Open Proposals, Risks & Blockers

    CategoryDescriptionImpactResolverNext Step
    BlockerDatabase migration delayBlocks sprint progressMarkIdentify root cause

    Section 6 — Ready-to-Send Follow-Up Recap

    Hi team, following our sprint planning: we are moving forward with REST for the API and have deferred the UI overhaul to focus on stability. Dave is handling the API documentation (due Friday). Mark is currently blocked by the database migration, which we need to address immediately to maintain our timeline. Please reach out if you have updates on the migration status. [72 words]

    Accountability & Readiness Verdict: READY

    Next steps

    • Copy the Recap into the project Slack channel.
    • Schedule a 15-minute sync with Mark regarding the database blocker.
    • Log A1 into the Jira sprint board.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Turn Messy Meeting Transcripts & Chat Threads into Accountable Action Items

    Generic AI meeting summaries blend casual brainstorming with actual decisions, invent deadlines nobody agreed to, and bury critical blockers in paragraphs of text.

    Meeting Action Extractor processes raw call transcripts, voice-note dumps, and Slack/Telegram threads with the rigor of a Technical Program Manager — separating approved decisions from open proposals and flagging every unassigned task.

    What It Delivers (Complete 6-Section Operations Blueprint)

    1. Meeting Metadata & Signal Snapshot: Compact table showing participants detected, finalized decisions count, assigned vs. unassigned action items, and open blockers.
    2. 3-Sentence Executive TL;DR: Strictly 45 to 80 words covering why the meeting happened, the single biggest outcome, and the immediate critical path.
    3. Finalized Decisions Log: Bullet list of locked-in choices with stated rationale and exact speaker/timestamp attribution ([Speaker, 14:20]).
    4. Prioritized Action Items Tracker (P0 / P1 / P2): Structured Markdown table with Task ID (A1, A2...), Priority, Verb-First Action, Owner, Due Date, and Dependencies.
    5. Open Proposals, Risks & Blockers Table: Separates deferred ideas (Parking Lot) and technical/operational blockers from finalized decisions.
    6. Ready-to-Send Follow-Up Recap (60–120 Words): Copy-pasteable message formatted for Slack, Telegram, or Email to align the team immediately.

    Strict Attribution & Anti-Hallucination Guardrails

    • Zero Invented Owners or Deadlines: Never guesses task owners or calendar dates. Labels missing owners as Unassigned (Needs Owner) and missing deadlines as TBD (Needs Date).
    • Brainstorm vs. Decision Separation: Never turns rejected ideas or unapproved proposals into committed tasks.
    • Secret Redaction & 3-Tier Verdict: Automatically redacts leaked API keys/passwords in raw chat dumps and ends with READY (all tasks assigned and dated), NEEDS INPUT (lists unassigned/undated items to confirm), or HOLD (flags falsified minutes or leaked secrets).

    Quick Start Prompt

    "Use meeting-action-extractor to turn this 45-minute sprint planning transcript into decisions, a P0-P2 action table, and a Slack recap."

    Requires text transcripts, not audio. Cannot calculate calendar dates from relative terms unless meeting date is provided. Redacts secrets and flags leaks rather than processing them.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    5.0
    1 review
    5
    1
    4
    0
    3
    0
    2
    0
    1
    0

    This skill has been used 2 times.

    Trust & safety

    Security scan

    Security scan passed 18 days ago

    • Free with an Agensi account

    Permissions

    Read Files
    Write Files
    Listed1 month ago
    Updated18 days ago

    What's inside

    Frequently Asked Questions

    The skill requires a text-based transcript or chat export to function. While it cannot process raw audio files like MP3s, it is compatible with text outputs from tools like Zoom, Otter.ai, Slack, and Fireflies.

    Unlike generic prompts, this skill uses a specialized validation protocol that flags tasks with missing owners or deadlines as 'NEEDS INPUT.' It also enforces strict speaker attribution to prevent the AI from hallucinating who said what during the discussion.

    Yes, the skill is framework-agnostic and works with any text-based input. You can paste in disorganized notes, Slack threads, or even messy brainstorm logs, and it will apply the same structured synthesis to create a clean executive summary.

    You will receive the 'Meeting Action Extractor' skill configuration, including the Core Execution Protocol logic and the 6-section operations blueprint template. This allows you to immediately begin generating structured summaries and risk logs from your meeting data.

    To ensure accurate deadlines, you should include the current date in your input. The skill can identify a 'Friday' deadline, but it requires the meeting's context to assign that to a specific calendar date and year.

    New in Business & Operations