THE INTERNET IS THINKING
Earlier sources: AI coding agents
AI coding agents use models and development tools to inspect, modify, and test software. Follow coding workflows, repository tooling, evaluations, and evidence about where these systems succeed or fail.
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 ↗85 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
smixs/pi-durable-telegram ↗
Always-on durable pi agent on your server that you talk to in Telegram. Survives crashes and restarts, keeps history in SQLite. pi-durable Telegram gateway.
- Hacker News · Posted on Hacker News · First archived
Ask HN: Can we expect LLM-based coding agents to become noticeably faster? ↗
I am thinking primarily about "single-threaded" usage of Claude etc which I'm still using as if it was still 2025. I just don't find the mental context switching worth it.
- GitHub · Repository created · First archived
vvojciech/pitwall ↗
A pit wall for your coding agents, inside iTerm2. See every Claude Code and Codex session at once, Allow, Deny and answer questions without switching tabs, undo a turn, and watch your plan limits.
- Hacker News · Posted on Hacker News · First archived
Study on AI coding agents finds gains "absorbed" by human review "bottleneck." ↗
Hacker News discussion.
- GitHub · Repository created · First archived
ArcfluxResearch/dudes ↗
Run Claude Code, Codex and Pi agents toward goals the daemon verifies itself. Give each agent a role and a goal with a completion check; Dudes keeps it working until its own run of the check passes. Rust daemon and CLI, with a 3D office view in the browser.
- GitHub · Repository created · First archived
kauakdev/kauak ↗
A Sims-style isometric office that shows what your Herdr coding agents are doing, live
- arXiv · Published · First archived
Returning to Development States: When Full Sandbox Snapshots Help Coding Agents ↗
Coding agents often make useful progress before they fail, but retry-from-scratch discards the machine state created along the way. This makes failure recovery an interesting search problem: if intermediate environment states preserve meaningful progress, restoring and branching from them could outperform repeatedly restarting from the initial state. We implement Go-Explore-style search over full Daytona snapshots and find that higher and more efficient solve rates depend mainly on progress that sits outside the repository, such as databases, services, and other runtime state. Our results sugg…
- Hacker News · Posted on Hacker News · First archived
Show HN: CircleIt – Circle something on a web page, your coding agent fixes it ↗
Hi HN! Agents have made our development so much faster and we've found that the bottleneck has moved to feedback from the rest of the team. We have marketers and sales teams that were screenshotting pages and providing notes via Google Docs which then had to be translated to prompts by devs. So we built CircleIt, a Chrome extension plus a Claude Code/Codex plugin. Anyone on the team can circle (and more!) things on any page (live/staging/dev), ex…
- GitHub · Repository created · First archived
abacusai/simless ↗
Simulator-free, headless SwiftUI checks for AI coding agents: render screens as accessibility trees and hot-reload views in ~2 s, without the iOS Simulator.
- GitHub · Repository created · First archived
NazmusSayad/offsite ↗
Claude Code in your browser. Shares sessions with the CLI, so you can start in one and resume in the other. Includes a terminal, file previews, and git diffs.
Live window recorded . This is the ingestion observation time, not the publication or discovery date of each source.
How to assess evidence about AI coding 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.