Current thesis
Use `radar.md` for the live thesis, changed thesis, and open questions that deserve attention now.
Python CLI · Markdown repo
Track agent signals in plain Markdown
Keep a versionable Markdown radar of agent product moves, field notes, and storage implications—stdlib CLI only.
$ python scripts/agent_radar.py brief
CORE FILES
-
radar.md -
agent-watchlist.md -
user-field-notes.md -
playbook.md -
storage-angle.md
radar.md Thesis: local agents win on evidence · Changed: sandbox snapshot noise · Open: who owns artifact storage
Markdown-first · stdlib CLIdaily append · weekly synthesis
Markdown-first radar
Signals you can edit and version
Agent Radar keeps thesis, watchlist, field notes, playbooks, and storage implications in plain Markdown files — plus a stdlib Python CLI to scaffold, validate, and brief the repo.
Editable thesis radar.md · current week
Current thesis Local agents win on evidence
Coding agents that keep workspace artifacts local are easier to audit than chat-only hosted flows.
Changed Sandbox + snapshot noise
More products advertise sandboxes; few document snapshot retention or replay cost.
Open questions Who owns the artifact store?
When agents write logs, checkpoints, and playbooks, which storage path is intentional vs incidental?
Core Markdown files
A simple structure for high-signal tracking
Each file has a clear job. Daily and weekly folders append signals; automation and cloud-agent workflows keep the cadence without turning the repo into a platform.
- 01
Thesis
`radar.md` holds current thesis, changed thesis, and open questions.
- 02
Watchlist
`agent-watchlist.md` tracks mainstream and emerging agents with evidence labels.
- 03
Field notes
`user-field-notes.md` and `playbook.md` capture real workflows and reusable patterns.
- 04
Storage angle
`storage-angle.md` records workspace, sandbox, artifact, log, and replay implications.
In the terminal
In the terminal
Agent watchlist
Track mainstream and emerging agents in `agent-watchlist.md` — product moves, not launch hype.
Real user evidence
Capture workflows, complaints, tricks, and failure cases in `user-field-notes.md` with source discipline.
Reusable playbooks
Promote patterns that generalize beyond one incident into `playbook.md` — prompts, setup tricks, recovery paths.
FAQ
What is Agent Radar not?
Not a news dump, crawler framework, vector store, or complex knowledge base. It is a lightweight editable Markdown structure with a small Python CLI.
How does the CLI work?
Python 3.10+ and only the standard library. `init`, `daily`, `weekly`, `monthly`, `status`, `validate`, and `brief` scaffold and check the repo from any subdirectory.
Can it run on a schedule?
Yes. GitHub Actions workflows support scheduled cloud-agent runs with source lanes, scoring, telemetry, and bilingual report output.
What about private sources?
Authorized logged-in or private signals may inform the radar, but public output must be anonymized and public-safe. See `docs/maintenance.md` and `SECURITY.md`.