THE INTERNET IS THINKING
Earlier sources: AI agents
AI agents use models to choose actions, call tools, and work toward a goal. This thread follows agent frameworks, memory, orchestration, evaluation, and safeguards; autonomy and reliability vary between systems.
SYMTRI organizes observations from public sources into a living map. Connections reflect shared subject matter; they do not establish causation or endorse a source.
Open in the map ↗845 classified signals in the fourteen-day window. These counts include signals awaiting readable evidence.
Earlier sources
Up to ten sources with readable evidence, ordered by their feed timestamp. These archive sources can be older than the live fourteen-day window. This list is a sample, not a measure of total activity. Dates distinguish paper publication, repository creation, model/dataset activity, Hacker News posts, and KEV catalog additions. “First archived” records when SYMTRI first stored the signal.
- GitHub · Repository created · First archived
ethanplusai/astra-flash-orchestrator ↗
Astra plans and reviews; DeepSeek Flash builds. A native Codex workflow with phased tasks, verification, safe installation and reversible setup.
- GitHub · Repository created · First archived
ai-4Work/iWork ↗
自主办公智能体
- GitHub · Repository created · First archived
Blackman99/deskfolk ↗
The AI teammates you'd want, on your own Mac: persistent bots in chats and groups, a split-pane desktop workbench with terminals, a live flow board for every job, per-turn model choice that says why, and experimental remote access. Bring your own model endpoints and MCP tools. Alpha.
- GitHub · Repository created · First archived
FourBis/4bis.relay-showcase ↗
Experimental local AI workspace for multi-repository conversations, persistent tasks and reviewable pull requests. Built by FourBis. MIT.
- GitHub · Repository created · First archived
ArtemPavlov1994/polymarket-prediction-bot ↗
Polymarket trading bot for prediction markets — browse CLOB markets, watch the order book in the terminal, run edge detection, liquidity provision and cross-market arbitrage strategies with paper trading and risk limits. Educational open-source toolkit — not financial advice. Unofficial community project, not affiliated with Polymarket.
- runtimewire.com · Source date · First archived
Anthropic’s Project Parka sits through meetings and assigns Claude agents the homework ↗
Anthropic is developing a meeting recorder inside Claude Desktop that can turn action items into tasks for Claude Cowork and Claude Code, according to application interfaces reviewed by RuntimeWire. The project is internally codenamed Parka. A Claude Desktop macOS package carrying August 18th, 2026 filesystem timestamps contains a detailed contract for recording meetings, streaming speaker-attributed transcripts, generating notes and summaries, and extracting structured follow-up work. Public packaged builds still mark the feature as unavailable, and Parka may never be its public name. Read th…
- runtimewire.com · Source date · First archived
Cursor built a hidden terminal command for steering its desktop AI agents ↗
The feature is called Desktop Bridge. It adds a hidden cursor desktop command with two operations: cursor desktop ls cursor desktop send [text...] Read the full story on RuntimeWire
- runtimewire.com · Source date · First archived
Kimi Work secretly attaches raw records from five recent agent sessions to feedback reports ↗
Moonshot AI's Kimi Work desktop app is designed to attach raw diagnostic records from five recent agent conversations whenever a user submits feedback, a RuntimeWire analysis of the Windows application found. The collection reaches beyond the task a user is reporting. Kimi Desktop sorts every locally registered Work conversation by its last update time, takes the newest five and asks a bundled background service to create a separate ZIP archive for each one. It uploads those archives alongside a desktop diagnostic-log package before sending the feedback text. Read the full story on RuntimeWire
- runtimewire.com · Source date · First archived
ChatGPT Wallet: OpenAI Is Building Agentic Payments ↗
OpenAI is preparing a payment product called ChatGPT Wallet that would let users save cards for ChatGPT to use while carrying out tasks, according to interface code RuntimeWire found in the Windows ChatGPT/Codex desktop application bundle . The application bundle contains an announcement modal for the wallet, a secure card-entry session and separate states for saving a card, retrying after a failure and confirming success. The accompanying copy tells users that ChatGPT can use saved payment methods to carry out tasks. Read the full story on RuntimeWire
Live window recorded . This is the ingestion observation time, not the publication or discovery date of each source.
How to assess evidence about AI agents
Check each original source for its methods, release status, limitations, and supporting data. A repository announcement, a preprint, and a published study provide different kinds of evidence. SYMTRI’s subject classification helps discovery; it does not independently validate the source’s claims.
See the collection and counting methodology ↗ · Read the weekly synthesis ↗
Related threads
Explore related regions
These navigation suggestions are part of the map’s editorial structure, rather than measured evidence of a connection.