sorted / seed
CLAUDE.md0 starsChanged 45 days agoMITView on GitHub
The demo's photos, labels, map layers and simulated data. Some of it is source and some is generated; never hand-edit the generated files.
In this repository11 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- content/CLAUDE.mdCLAUDE.md
- evals/CLAUDE.mdCLAUDE.md
- scripts/CLAUDE.mdCLAUDE.md
- static/CLAUDE.mdCLAUDE.md
- checkSkill
- evalSkill
- rebrandSkill
- resetSkill
- screenshotsSkill
The file
We split this CLAUDE.md into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Everything here is fictional or credited: No real report, no real person, and no real council's name or crest.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
seed/
The demo's photos, labels, map layers and simulated data. Some of it is source and some is generated; never hand-edit the generated files.
| Path | What it is | Made by |
photos/case_NNN_before.jpg, _after.jpg | AI-generated photos of each case, before and after clearing | Committed |
cases/labels.json | What each photo shows, set by a person looking at it | Committed, by hand |
live_demo/ | The five demo-kit photos, with GPS written in, and LIVE_DEMO.md | scripts/build_kit.py |
*.geojson | Streets, wards, boundaries and railway and National Highways land | scripts/build_geo.py, from sources in NOTICE.md |
demo_incidents.json, history.json, model_cache.json, photos_web/ | The simulated reports and cached readings, dated relative to today | scripts/build_seed.py, not in git |
cases/labels.jsonis ground truth twice over. The seed builder turns it into each case's reading, andevals/score.pyscores the model against it. Change a label only by looking at the photo, never to move a score. A label change moves the eval results, so re-score and say so indocs/EVALS.md. The edit hook blocks agents from changing it.- A photo's bytes are its identity. The cache and the evals both key photos by SHA-256. Re-saving a photo, even losslessly, breaks its cached reading and means the shipped prompt is no longer measured on it (
tests/test_units.pyfails). Rebuild the seed and re-run the evals after changing any photo. - The kit photos and the simulated places move together. The GPS in each
live_demo/photo, the hotspots and the merge target all come fromscripts/demo_places.py. After changing a place, runbuild_kit.py, thenbuild_seed.py. - Everything here is fictional or credited. No real report, no real person, and no real council's name or crest. The licence for each map layer is in
NOTICE.md; add a row there for any new file.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.