Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Decision Pulse: AI Cognitive Load Dashboard

    1

    See when a day of AI work is overloading your brain: a one-page dashboard with gentle, research-backed suggestions.

    Free

    1 useSecurity scanned
    Decision Pulse: AI Cognitive Load Dashboard

    Decision Pulse: AI Cognitive Load Dashboard

    Example session with this skill installed

    How's my AI usage today? Am I overloaded? Should I take a break?

    • Read your context and instructions
    • Compiled the decision pulse

    Summary

    That day: 6 conversations, 216.4 min in conversation; busiest hour 16:00–17:00 (18 messages); longest stretch without a break 86.2 min. Four stages: decision density

    low, recovery window

    medium, fatigue signs

    medium, judgement quality

    low.

    Dashboard saved as decision-pulse-2026-10-05.html (role: Operations / HR, interface: English).

    Four stages (each levelled on its own, never summed)

    StageValueThresholds (low / medium / high)Level
    ① Decision density23.8 messages per active hour25 / 45 (rule of thumb, HR preset)Low
    ② Recovery window86.2 min longest stretch without a 10-min break60 / 120 (10 min and 6 h are research-based)Medium
    ③ Cognitive fatigue9.7 conversation switches per active hour6 / 12 (rule of thumb)Medium
    ④ Judgement quality7.7% of long AI replies answered within seconds10% / 25% (rule of thumb)Low

    A few things worth noticing today

    Recovery window (medium)
    You've been going for quite a while without a pause. Before the next thing, maybe get up, walk around, look out the window, and give yourself ten minutes away from the screen.
    Why: breaks of ten minutes or more are what actually help you recover; looking into the distance or at something green helps your attention come back.
    <sub>Albulescu et al. 2022 (confidence: medium-high); attention restoration theory, Kaplan 1995 (confidence: medium)</sub>

    Cognitive fatigue (medium)
    Before you jump to another conversation, take half a minute to jot down where you are and what comes next. It's easier to pick up again, and easier to let go when you leave.
    Why: writing a plan for picking back up keeps your attention from getting stuck on the last task.
    <sub>Ready-to-resume plans, Leroy & Glomb 2018 (confidence: medium-high)</sub>

    Details (excerpt)

    Density: 86 messages you sent, 1,782 characters typed vs 82,573 characters written by the AI; 6.97M tokens (410K thinking); the AI stopped to ask you 12 times; 13 high-stakes messages (7 information spikes, 9 in a fatigue context)

    • Hourly: quiet morning (3 msgs/hour), then 15:00–18:00 carried most of the load (17 / 18 / 16 / 13)

    Recovery: 216.4 active minutes, 7 real breaks of 10+ minutes, longest break 83.8 min; never crossed the 6-hour sustained-control line

    Fatigue: 35 switches, 30 of them jumping back to an earlier conversation; up to 3 conversations open at once (15:13); sharpest 15 minutes 17:50–18:05 with 6 switches; no late-night messages

    Judgement: 59 replies, median 58 s before replying; 4 of 52 long AI replies answered within seconds; 26 retries after tool errors; 61 explore vs 84 produce calls

    Next steps

    • Take a 10-minute screen break before the next work block
    • Before switching conversations late in the afternoon, jot a one-line "where I am / what's next" note

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

    About this skill

    AI adds a lot of small decisions to a day. The agent finishes a step and stops to ask you. You keep several conversations open and hop between them. You sit for hours. None of it is hard on its own, but it adds up, and you rarely notice when you're tired.

    Decision Pulse reads one day of your Claude Code and Codex transcripts, locally on your own computer, and turns it into a one-page dashboard that shows where the day got heavy.

    How it reads your day

    It follows a mechanism chain where each stage pushes pressure into the next:

    1. Decision density: how many things needed your call (messages per active hour)
    2. Recovery window: whether you actually took a break (longest stretch without a 10-minute break)
    3. Cognitive fatigue: how fragmented your attention was (conversation switches per active hour)
    4. Judgement quality: whether you still read carefully (share of long AI replies you answered within seconds)

    Each stage gets its own low / medium / high level. They are never summed into a single score, so you can see which stage to deal with first.

    What you get

    • A one-line summary you can read without opening the page
    • A "right now" panel when you check today: how long the current stretch has lasted
    • Four colour-coded stage cards with values and the thresholds behind them
    • A few suggestions for stages at medium or high, written like a colleague's reminder and each tied to a cognitive-science mechanism and paper
    • Details: conversations, characters typed and read, tokens, high-stakes moments, morning/afternoon/evening split, work and break timeline, switching, prompt length and review behaviour
    • Role presets (engineering, ops/HR, sales, design, manager, general) that adjust what counts as normal density, while break, switching and judgement thresholds stay the same for everyone
    • An optional desktop clock (Tkinter) with four arcs showing the stage levels, refreshed every 30 minutes, with no notifications
    • Full English and Traditional Chinese interface

    Research basis

    Thresholds cite Albulescu et al. 2022 (breaks), Blain et al. 2016 (sustained control), Leroy 2009 and Leroy & Glomb 2018 (attention residue), Mark et al. 2008 (interruptions), Parasuraman & Manzey 2010 (automation complacency) and others, each with a stated confidence level. Thresholds without research behind them are labelled "rule of thumb". Ego depletion, the 90-minute ultradian rhythm and Pomodoro are deliberately not used.

    Privacy

    • Reads only local transcripts in ~/.claude/projects and ~/.codex/sessions
    • The script makes no network calls and uploads nothing
    • Daily results stay in the skill's own history/ folder on your machine

    What it doesn't do

    It doesn't judge whether your decisions were right, compare days, push reminders, give a single overall score or guess your emotions. Suggestions are fixed texts triggered by rules, not generated on the fly.

    Created and maintained by Wesley Cheng, author of github.com/chengwesley/decision-pulse (MIT).


    繁體中文說明

    Decision Pulse 讀取你電腦上當天的 Claude Code 與 Codex 對話紀錄,依「決策密度 → 恢復窗口 → 認知疲勞 → 判斷品質」這條機制鏈產出一頁儀表板。四個階段各自分成低/中/高,不加總成單一分數;中或高的階段會給出有認知科學依據、像同事提醒般的建議。全程在本機執行、不連網、不上傳。介面完整支援繁體中文與英文,也可依職能(工程、營運・HR、業務、設計、主管)調整決策密度門檻。

    Looks at one day at a time: no cross-day comparison or baseline. Does not judge whether your decisions were right, only the conditions you decided under. Codex transcripts lack AI reply length, tokens and reply timing, so those metrics cover Claude Code only (the page marks this). Density and switching thresholds are rules of thumb that may need calibrating. The design spec (references/spec.md) and config comments are in Chinese; the dashboard, suggestions and desktop clock are fully in English or Traditional Chinese.

    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

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    1 use

    Times used to date

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 1 month ago

    • Free with an Agensi account

    Requires

    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    File Scopes

    ~/.claude/projects/**
    ~/.codex/sessions/**

    Terminal: runs the bundled Python script (scripts/scan_today.py). Read Files: reads only your local Claude Code transcripts (~/.claude/projects) and Codex sessions (~/.codex/sessions) for the chosen day. Write Files: writes the dashboard HTML, that day's JSON and your config.local.json inside the skill folder (history/). No network access: the script makes no network calls and uploads nothing. The generated HTML page loads fonts from Google Fonts when you open it in a browser.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    Decision Pulse quantifies 'hidden' AI fatigue by analyzing local Claude Code and Codex logs for decision density, attention fragmentation, and recovery gaps that often lead to burnout.

    The skill currently supports Claude Code and Codex; it reads local transcript files located in your machine's project and session directories.

    No, the skill is designed for privacy and operates entirely offline by reading local logs, making zero network calls or external data uploads.

    The dashboard includes specialized presets for Engineering, Ops/HR, Sales, Design, and Management to ensure 'normal' activity levels are calibrated to your specific workload.

    You receive the full local dashboard application, the optional Tkinter desktop status clock, research-backed suggestion modules, and dual-language support (English and Traditional Chinese).

    Updates are released periodically to reflect new cognitive science research or to support updates in Claude Code and Codex file structures; you can download these via the marketplace.

    New in Productivity