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 ↗86 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.
- arXiv · Published · First archived
Compact Documentation for Coding Agents: A Benchmark, an Optimizer, and Why It Does Not Transfer ↗
We investigate whether natural-language documentation helps coding agents resolve software issues, and we build the tools to construct and evaluate it. We introduce a roundtrip benchmark that scores code descriptions by whether code regenerated from them passes the original tests, and show that completeness, not length, drives a description's fidelity. Using the benchmark as an optimization signal, we discover a description-writing prompt that reaches full fidelity and generalizes to unseen files. We then test the hypothesis that motivated the work: that better documentation helps an agent res…
- Hacker News · Posted on Hacker News · First archived
Cargo-atlas – A compiler-accurate Rust code map for AI coding agents ↗
Hacker News discussion.
- GitHub · Repository created · First archived
gvergnaud/bise ↗
a multi-agent harness, made for humans :* meet your team lead. you stay in flow, i run the agents.
- GitHub · Repository created · First archived
kiselas/every-frame-is-code ↗
Every frame is code: an Agent Skill for Claude Code, Codex and ChatGPT that makes motion graphics, 3D scenes and games in HTML/JS and renders them to MP4 frame by frame. Canvas 2D, WebGL, Three.js, GLSL, Web Audio.
- GitHub · Repository created · First archived
magik-ai/murmur ↗
Open-source orchestrator for coding agents: run Claude Code and Codex as one team on a machine you own.
- GitHub · Repository created · First archived
HimalayanNomads/drawa ↗
A browser canvas for coding agents: run Claude Code, OpenCode and Codex sessions side by side, watch what they touch, review every change.
- GitHub · Repository created · First archived
matank001/clodfarm ↗
clodfarm (say it out loud): a farm of Claude Code agents. Plant a mission, they split it into sub-agents, open the work, and pace themselves on each account's real 5-hour and weekly usage. Steer it from the Claude app.
- GitHub · Repository created · First archived
leciric/agentbox ↗
Devin-style that run on your own machine: each agent gets its own Linux machine, desktop and branch, directed by a lead chat. Claude Code, Codex and OpenCode.
- arXiv · Published · First archived
Coding Agents for Generalized Task and Motion Planning Problems ↗
Task and motion planning (TAMP) problems remain difficult even with full observability and object-centric states because discrete decisions are tightly coupled to geometric, kinematic, and dynamic constraints. Generalized TAMP addresses this difficulty by exploiting regularities across problem instances to reduce planning effort on new instances. However, existing methods require substantial TAMP-specific engineering. We investigate whether coding agents can automate this process by synthesizing programs that generalize across instances. Given a task description and simulator access, each agen…
- GitHub · Repository created · First archived
korengast/angelia ↗
The coding agent CLI you already trust becomes your personal assistant, on WhatsApp and Telegram. Runs on your machine.
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.