惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

C
Cisco Blogs
罗磊的独立博客
D
Docker
Microsoft Azure Blog
Microsoft Azure Blog
C
CXSECURITY Database RSS Feed - CXSecurity.com
V
V2EX
Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
Cisco Talos Blog
Cisco Talos Blog
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
I
InfoQ
S
Securelist
K
Kaspersky official blog
博客园 - 司徒正美
爱范儿
爱范儿
Scott Helme
Scott Helme
B
Blog RSS Feed
H
Help Net Security
博客园 - 聂微东
Hugging Face - Blog
Hugging Face - Blog
Stack Overflow Blog
Stack Overflow Blog
AI
AI
Blog — PlanetScale
Blog — PlanetScale
Webroot Blog
Webroot Blog
P
Proofpoint News Feed
V
Visual Studio Blog
Cyberwarzone
Cyberwarzone
P
Privacy International News Feed
M
MIT News - Artificial intelligence
Google DeepMind News
Google DeepMind News
酷 壳 – CoolShell
酷 壳 – CoolShell
博客园 - Franky
IT之家
IT之家
云风的 BLOG
云风的 BLOG
MyScale Blog
MyScale Blog
L
LINUX DO - 热门话题
P
Palo Alto Networks Blog
S
Security Affairs
T
Threat Research - Cisco Blogs
S
Security @ Cisco Blogs
The Register - Security
The Register - Security
F
Full Disclosure
A
Arctic Wolf
C
Check Point Blog
Recent Announcements
Recent Announcements
P
Proofpoint News Feed
人人都是产品经理
人人都是产品经理
T
Tor Project blog
Latest news
Latest news
Schneier on Security
Schneier on Security
H
Hackread – Cybersecurity News, Data Breaches, AI and More

Show HN

GitHub - steveking-gh/firmion: Firmion is DSL and engine for firmware image generation. GitHub - villagesql/villagesql-skills: Agent skills for VillageSQL - gemini-cli-extension; claude-code-plugin GitHub - flightdeckhq/flightdeck: Observability and control plane for AI agents. CSP Radar GitHub - Light-Heart-Labs/DreamServer: Turn your PC, Mac, or Linux box into an AI server. LLM inference, chat UI, voice, agents, workflows, RAG, and image generation. GitHub - Diplomat-ai/diplomat-agent-ts: What can your TypeScript AI agent do to the real world? Scan your code. See which tool calls have zero checks Code Block Selector - Visual Studio Marketplace Prometheus dependency graph — interactive showcase | Riftmap Show HN: I made a vi-like modal keyboard plugin for Figma GitHub - run-llama/liteparse: A fast, helpful, and open-source document parser GitHub - dalemyers/Roar: A macOS CLI tool for notifications GitHub - district-solutions/open-agent-tools-coder: Enables small-to-large self-hosted ai models to use local source code when running tool-calling agentic workloads. We actively data mine 20,900+ (2+ TB) popular github repos using large and small ai models to create reuseable: json, markdown and parquet files for local-first tool-calling models. GitHub - progapandist/stripeek: A local TUI proxy for real-time Stripe API debugging, built for navigating complex payloads fast. GitHub - sir1st/hermes-desktop: All-in-one cross-platform desktop app for Hermes Agent — bundles Python + hermes-agent + hermes-web-ui GitHub - astefanutti/shaderbang: Shebang for Shaders Show HN: Generate Claude Code Workflows using Spec Driven Development approach GitHub - nixys/nxs-universal-chart: The Helm chart you can use to install any of your applications into Kubernetes/OpenShift Show HN: AI agents for UK GDAD PCF roles and their skills The Two Pillars: Mixer Mode and Meta-Software in the Reorganization of Software Work After AI GitHub - JaiCode08/teleport-env What 1,000+ Harness Experiments Taught Me About Self-Improving Agents Show HN: Liiists, a Markdown-first, iOS and CLI list app SwiperTab – Get this Extension for 🦊 Firefox (en-US) GitHub - kouhxp/fftext: Summarize, explain, fact-check, or translate any text, URL, or file. No GPU. No cloud. One command GitHub - sweetpad-dev/sweetpad: Develop Swift/iOS projects using VSCode GitHub - dogmaticdev/IRON: IRON a.k.a. Intermediate Representation Object Notation is a Interpreter/Database that is used to create Programming Languages. GitHub - sjhalani7/vaen: Package your AI coding harness into a portable .agent file, and share it across repos, teams, & the community without ever having to copy-paste instructions, skills, MCP config, or secrets. Show HN: Gandalf the Grader Show HN: Citadeld – replay any CI failure locally from a single file GitHub - tdortman/cuSBF: High-Performance GPU Super Bloom Filter coral-ai/claude-code-token-xray at main · Coral-Bricks-AI/coral-ai GitHub - ulyssestenn/funes: Funes is a Git-based framework for LLM-managed knowledge work: an AI Librarian ingests raw sources, builds an interlinked Markdown knowledge base, and uses it to produce cited reports, analyses, and other outputs. GitHub - ThatXliner/gah: Git Add Hunk, built for agents to use GitHub - harmont-dev/harmont-cli: Command-line client for the Harmont CI platform GitHub - brooksmcmillin/mcp-authflow: OAuth 2.0 Authorization Server framework for MCP servers GitHub - javaid-codes/audit-supply-chain-agents GitHub - amorey/gochan: A small library of common channel architectures for Go, inspired by Rust GitHub - arifozgun/OpenGem: Free, Open-Source AI API Gateway with Gemini, OpenAI & Anthropic Compatibility in 1 file GitHub - Pranesh950/BioPetals: 🌸 Run BIOxAI models at home, BitTorrent-style. Fine-tuning and inference up to 10x faster than offloading GitHub - cnguyen14/bounty-doctor: Diagnose a GitHub bounty issue before you waste hours: detects honeypot scam repos, AI-bot attempt swarms, and stale contests. Show HN: CoreMCP – MCP Server for On-Prem DBs Show HN: KittyHTML – Render HTML/CSS as an inline image in your terminal GitHub - bingud/filemat: Web-based file manager Show HN: TruthLens – Free multi-signal deepfake image detector GitHub - apexlocal-jz/claude-usage-tray: Windows system-tray app showing your Claude Code rate-limit usage at a glance. Zero deps, ~300 lines of PowerShell. Cross-IDE (works regardless of VS Code, Cursor, plain terminal). Release v0.1.2.1 · kouhxp/yapsnap GitHub - noopolis/moltnet: Self-hostable chat network for AI agents. Pre-built bridges for Claude Code, Codex, and the Claws. Rooms, DMs, history. No Slack bots, no Matrix, no glue code. GitHub - tamerh/enju: Coordinating Humans, AI Agents, and Compute as Peers on a Shared Workflow Graph Show HN: Continuity-auth – Respect-weighted rate limits for the open web GitHub - luml-ai/luml: AI lifecycle platform where engineers and agents track experiments, train models, and ship to production. GitHub - mrdanielcasper/CoreTex: A UNIX-inspired, biomimetic, flat-file AI harness and knowledge engine. GitHub - clemg/pierre-github: Pierre's diffs.com and trees.software for Github GitHub - lyriks-io/unspaghettit: Behavior-driven AI development without prompt spaghetti. GitHub - sofumel/claude-handoff-revive: Resume Claude Code work after rate/usage/context limits without replaying the prior transcript. Auto-saves at 90%/95% usage. Plugin-installable, 10 languages. GitHub - dotexorg/saferpc: Typed, end-to-end encrypted RPC over any bidirectional channel. GitHub - BeeZeeAgent/beezee: Agent harness orchestration Legato Next.js Boilerplate for Internal Tools · CoreUI GitHub - clark-labs-inc/clark-hash: Clark Hash, 32x smaller searchable sketches for embeddings GitHub - ZeroPointRepo/youtube-mcp: The fastest YouTube transcript + YouTube search MCP for AI agents. Try for free. Typing Mastery — climb toward 100+ WPM, deliberately GitHub - Andebugulin/Awareen GitHub - fayzan123/claude-workflow-composer: Visual desktop app for composing multi-agent coding workflows. Drag agents, attach skills and MCPs, wire handoffs, export to .claude/ GitHub - harshaneel/humanize: Best static AI text humanizer. Two research-grounded skills that work in any LLM (Claude, ChatGPT, Gemini, Codex): humanize beats perplexity-based detectors, ai-check produces forensic scoring with evidence-quoted flags. Nine levers, 50+ peer-reviewed sources, 2024-2026 detection literature. GitHub - StackOneHQ/stack-nudge GitHub - nodes-app/swift-markdown-engine: A native AppKit Markdown editor for macOS, built on TextKit 2 and bridged to SwiftUI. We hardened an LLM agent. Each defense we added made it more exploitable. GitHub - alkait/WhatsKept: Agent-queryable WhatsApp history from an iOS backup — a single Go binary. GitHub - octelium/cordium: Open-source, general-purpose sandbox platform for devs and AI agents that provides identity-based secure access to infrastructure without credentials. WAR.GOV/UFO Microfilm5 GitHub - scosman/videowright: Build animated explainer videos with your coding agent GitHub - dipankar/dscode: The code editor you can take apart. GitHub - zoharbabin/web-researcher-mcp: MCP server (Go) for AI assistants: web search, content extraction, academic/patent/news research. Multi-provider routing, 4-tier scraping, search lenses. Works with Claude, Cursor, and any MCP client. GitHub - ruvnet/RuView: π RuView turns commodity WiFi signals into real-time spatial intelligence, vital sign monitoring, and presence detection — all without a single pixel of video. GitHub - scanaislop/aislop: Catch the slop AI coding agents leave in your code: narrative comments, swallowed exceptions, as-any casts, dead code, oversized functions. 50+ rules across 7 languages (TypeScript, JavaScript, Python, Go, Rust, Ruby, PHP). Sub-second, deterministic, no LLM at runtime. MIT-licensed. GitHub - kouhxp/cheap-im: CPU-only voice agent approximating Thinking Machines' Interaction Models demo GitHub - unprovable/OrchidMantis: Orchid Mantis — standalone framework for Zero-Knowledge Proofs of eXploit (ZKPoX). GitHub - MarcellM01/TinySearch: Shrink the web for your local LLMs! GitHub - TangibleResearch/Halgorithem: A Algo designed to detect AI Hallucitions GitHub - DO-SAY-GO/freelang: I love freelang GitHub - CarpseDeam/Aura-IDE: An AI coding harness that shaped itself - Planner/Worker agents, repo awareness, surgical edits, validation, recovery, and safe diff approvals. GitHub - chojs23/concord: A feature-rich TUI client for Discord GitHub - tommyjepsen/awesome-ux-skills: UX & AI Product designs skills you can use today in Claude Code GitHub - aerf-spec/aerf: Agent Evidence Receipt Format (AERF) — an open specification for tamper-evident, independently verifiable records of AI agent actions. GitHub - kklimuk/docx-cli: CLI for AI agents (Claude, Codex) to read, edit, and comment on .docx files with full format fidelity. GitHub - Jwrede/tokentoll: Catch LLM cost changes in code review. Infracost for LLM spend. GitHub - samchon/ttsc: A `typescript-go` toolchain for compiler-powered plugins and type-safe execution + 500x faster lint integrated into compiler GitHub - Higangssh/homebutler: 🏠 Manage your homelab from chat. Single binary, zero dependencies. GitHub - olalie/tapmap: See where your computer connects and what stands out on a live world map. GitHub - Diplomat-ai/diplomat-agent: What can your AI agent do to the real world? Scan your code. See which tool calls have zero checks GitHub - Bajusz15/beacon: Open-source agent for secure remote access, monitoring, and deploys across home-lab and self-hosted machines like Raspberry Pi, N100, or any Linux server. Open web based TTY or tunnel Home Assistant and other local services securely without opening ports. BigTech AI News - Chrome 应用商店 GitHub - vinhnx/VTCode: VT Code is an open-source coding agent with LLM-native code understanding and robust shell safety. Supports multiple LLM providers with automatic failover and efficient context management. GitHub - michaelaz774/decision-engine: A decision operating system for startup founders, powered by Claude Code. Synthesizes wisdom from 25+ legendary founders and investors into interactive AI-driven decision frameworks. GitHub - Chrilleweb/dotenv-diff: Validate environment variable usage in your codebase GitHub - Lumen-Labs/brainapi2: BrainAPI is a knowledge graph–powered AI memory layer that transforms unstructured data into structured knowledge, enabling intelligent search, recommendations, and contextual memory for AI agents and applications. GitHub - familiar-software/familiar: Let AI watch you work. Familiar lets your AI update its memory, skills, and knowledge by watching your screen. GitHub - skorotkiewicz/rudo: A small, elegant dock for Wayland GitHub - muxshed/shed: One stream in, or many. Every destination, simultaneously. No cloud middleman, no per-channel fees, no limits. make sidebar/address bar rounded corner toggleable
Tokenyst - Visual Studio Marketplace
herrj · 2026-06-01 · via Show HN

Track your GitHub Copilot Chat and Copilot CLI usage, estimate token costs, and watch your combined spend against a monthly budget — all from a VS Code sidebar and the status bar.

Tokenyst is strictly local. It reads Copilot's own session files on your machine, estimates costs from a pricing table baked into the extension, and stores everything in a single JSON file under your home directory. Nothing is sent to any Tokenyst server.

Features

  • Usage Stats panel — a dedicated sidebar view with:
    • A monthly budget card showing spend vs. your cap
    • Today / This Week and Avg Daily / Weekly KPIs (with a period filter)
    • A pace projection for the current billing period
    • Collapsible breakdowns by day of week, model, token type (input / output / cache, with a cache-reuse %), source (Chat vs CLI), and repo — each foldable individually, with an Expand all / Collapse all toggle
  • Session breakdown — a sortable list of your individual Chat and CLI sessions, each with a readable title (its first prompt), so you can pinpoint your most expensive sessions. Sort by date, source, title, or your chosen metric (credits, input, or output tokens); filter by an independent date range; and click any session to expand a detail view (tokens, cache, model, repo)
  • Optimization insights — automatic, colour-coded tips based on your usage patterns: low cache reuse, model and repo hotspots, and expensive-but-low-output sessions
  • Two sources, one budget — tracks both Copilot Chat and the Copilot CLI, combined into a single monthly total (matching GitHub's shared usage-based billing), with a By source breakdown so you can see how much came from each
  • Status bar indicator — live $(graph) N cr of credits spent this period
  • Budget periods — calendar month by default, or anchored to your plan's renewal day
  • Real costs — Tokenyst uses the actual credit cost GitHub records (per-request credits for Chat, the CLI's reported "AI Credits" for the CLI), and shows the underlying token counts for transparency; it only falls back to a token-based estimate when no credit value was recorded
  • Historical import — backfill stats from your existing Copilot session history
  • Manual allocations — add custom allocations directly from the UI with credit amount, model name, and optional repository tracking, and remove them again from a picker list

How it works

Tokenyst reads from two local sources and combines them into one budget:

Copilot Chat

  1. The VS Code Copilot Chat extension records each chat session under <VS Code User>/workspaceStorage/<workspace>/chatSessions/.
  2. Tokenyst watches those files and aggregates usage per session and model, reading the real token counts (promptTokens/completionTokens) that Copilot records on each completed request. Where GitHub records a real credit value for a request, Tokenyst uses it directly.
  3. The input counts are context-inclusive (system prompt, tool definitions, attached files, and conversation history), so they reflect what Copilot actually sends — not just your typed message.

Copilot CLI

  1. The GitHub Copilot CLI records each session under ~/.copilot/session-state/<session>/events.jsonl.
  2. Tokenyst watches those logs and reads the real AI-credit cost the CLI records for each model (the same "AI Credits" the CLI prints when it exits), so CLI spend matches GitHub exactly. Token counts are kept for transparency.

For Chat requests where no credit value is recorded — and for older CLI logs that predate the credit field — Tokenyst estimates the cost from a token pricing table baked into the extension (matching GitHub's usage-based billing). All usage is stored as allocations in ~/.tokenyst/config.json.

Requirements

  • VS Code 1.85.0 or newer
  • GitHub Copilot — the Copilot Chat extension and/or the GitHub Copilot CLI (the copilot command, installed via npm install -g @github/copilot) must be installed and in use; that's what produces the session files Tokenyst reads. Either source alone is enough.

Remote development (dev containers, WSL, SSH, Codespaces)

Copilot's two data sources land on opposite sides of VS Code's client/remote split:

  • Copilot Chat sessions are persisted on the host (client) side — even for a dev container or other remote workspace. So Chat usage from your remote work is already in your host's storage.
  • Copilot CLI writes inside whatever environment it runs in — the container/remote if you use it there.

A single extension instance can only read one side. By default Tokenyst runs host-side (its extensionKind prefers ui), which tracks all your Copilot Chat — local and remote — plus Copilot CLI run on the host. You don't need to install it in the container for Chat tracking; it just works.

The one thing this default does not see is Copilot CLI run inside a container. If that's your workflow, tell VS Code to run Tokenyst in the remote instead by adding this to your User (or dev container) settings.json, then reload:

"remote.extensionKind": {
  "TokenystCopilot.tokenyst-copilot": ["workspace"]
}

VS Code will offer to install Tokenyst in the container; once it does, it tracks the container's Copilot CLI usage (and its data lives in the container's ~/.tokenyst/).

Note this is one or the other per setup: a container-side instance only sees the container's files, so in that window it tracks the container's Copilot CLI and none of your host-side usage — neither Copilot Chat (which is stored host-side) nor any Copilot CLI you run on the host. Each side keeps its own separate ~/.tokenyst/ data. In a dev container, Tokenyst shows a one-time hint pointing here so you can choose.

Getting started

  1. Install Tokenyst from the VS Code Marketplace.
  2. Open the Tokenyst view from the activity bar (look for the Tokenyst icon).
  3. Run Tokenyst: Enable Copilot Tracking from the Command Palette. Tokenyst will check for Copilot Chat and Copilot CLI session files and offer to import any existing history. A single toggle covers both sources.
  4. Run Tokenyst: Set Monthly Budget to pick a Copilot tier preset (Pro, Business, Pro+/Enterprise) or enter a custom amount.
  5. (Optional) Run Tokenyst: Set Renewal Date if your plan renews on a day other than the 1st.

Commands

All commands are available from the Command Palette under the Tokenyst category.

Command Description
Set Monthly Budget Choose a Copilot tier preset or enter a custom monthly cap
Set Renewal Date Day of month (1–31) your plan renews; blank uses the calendar month
Enable Copilot Tracking Start watching Copilot Chat and CLI session files; offers a historical import
Disable Copilot Tracking Stop tracking (both sources)
Add Manual Allocation Add a custom allocation with credit amount, model, and optional repository
Delete Manual Allocation Pick from a list of your manually-added allocations and remove one
Import Historical Usage Backfill allocations from the last 30 days of sessions
Force Sync Re-scan sessions and update usage immediately
Refresh Re-scan sessions and refresh the UI

Settings

Setting Default Description
tokenyst.syncIntervalSeconds 300 How often (in seconds) to scan Copilot session events in the background
tokenyst.showStatusBar true Show active budget usage in the status bar

Data & privacy

Tokenyst keeps everything on your machine:

  • ~/.tokenyst/config.json — your budget, renewal day, and recorded usage allocations
  • ~/.tokenyst/copilot.log — debug log (only written when debug logging is enabled)
  • Reads your VS Code Copilot Chat session files under workspaceStorage/*/chatSessions/ (written by the Copilot Chat extension)
  • Reads your Copilot CLI session logs under ~/.copilot/session-state/*/events.jsonl (written by the Copilot CLI)

Tokenyst makes no network requests and has no external dependencies — it reads only local Copilot session files and writes only to ~/.tokenyst/. Pricing data is baked into the extension; no pricing or billing service is contacted.

Editing allocations by hand

Normally you remove a manual allocation with Tokenyst: Delete Manual Allocation (a picker lists every manually-added entry). If you'd rather edit the data directly — for example to fix a typo or bulk-remove entries — you can edit ~/.tokenyst/config.json yourself:

  1. Close VS Code (or at least the Tokenyst view) so the file isn't rewritten under you.

  2. Open ~/.tokenyst/config.json. On Windows this is C:\Users\<you>\.tokenyst\config.json; on macOS/Linux it's ~/.tokenyst/config.json.

  3. Find the allocations array. Manually-added entries are the ones with "manual": true and/or an "externalId" starting with "manual-":

    {
      "costUsd": 0.5,
      "model": "copilot-gpt-4",
      "inputTokens": null,
      "outputTokens": null,
      "cacheCreationTokens": null,
      "cacheReadTokens": null,
      "filesModified": [],
      "at": "2026-06-01T12:42:46.000Z",
      "provider": "copilot",
      "externalId": "manual-1717245766000-a1b2c",
      "manual": true
    }
    

    Older manual entries created before tagging existed may have neither marker — they're identifiable as "provider": "copilot" entries with all four token fields (inputTokens, outputTokens, cacheCreationTokens, cacheReadTokens) set to null. Tokenyst treats those as manual too.

  4. Delete the entire object (including its surrounding { } and the trailing comma) from the array, then save. Make sure the file is still valid JSON.

  5. Reopen VS Code; the Usage Stats panel will reflect the change on the next refresh.

Only delete entries you recognize as manual. Synced Copilot allocations always carry numeric token counts and a copilot-chat-… (Chat) or copilot-cli-… (CLI) externalId; removing those just makes them reappear on the next sync.

Note: Spend is shown in credits to match Copilot's usage-based billing (100 credits = $1). Where Copilot records a real credit value for a request, Tokenyst uses it directly; otherwise it estimates from token counts and a built-in pricing table (with a cache discount for the repeated system prompt and tool definitions), so those figures are approximate.

Please send any questions or comments to contact@tokenyst.dev

Note: Are you an engineering manager or team lead? I'm exploring a team dashboard and other QOL features, please contact if interested.

License

MIT