diff --git a/.agents/skills b/.agents/skills
new file mode 120000
index 0000000..42c5394
--- /dev/null
+++ b/.agents/skills
@@ -0,0 +1 @@
+../skills
\ No newline at end of file
diff --git a/.claude/skills b/.claude/skills
new file mode 120000
index 0000000..42c5394
--- /dev/null
+++ b/.claude/skills
@@ -0,0 +1 @@
+../skills
\ No newline at end of file
diff --git a/.gitignore b/.gitignore
index df92e95..54d6f64 100644
--- a/.gitignore
+++ b/.gitignore
@@ -24,9 +24,13 @@ docs/_build/*
 coverage.xml
 htmlcov/*
 
+# IDEs and editors
 .idea/
 .vscode/
-scripts/download/data/*
+.zed/
+# ZED spellchecker
+codebook.toml
+
 *.egg-info
 
 # Version
diff --git a/AGENTS.md b/AGENTS.md
new file mode 100644
index 0000000..03c2a94
--- /dev/null
+++ b/AGENTS.md
@@ -0,0 +1,42 @@
+# AGENTS.md
+
+Guide for LLM coding agents working with the snowexsql repository.
+
+## When using the package
+
+DO NOT explore the codebase at first.
+Use the preferred pattern from the [README.md](README.md).
+Only use reverse engineering based on the code as a last resort.
+
+## When writing new code
+
+### Core Technologies
+
+Do not add new dependencies without approval from the team.
+Prefer existing core technologies used in this project:
+
+- **SQLAlchemy** + **GeoAlchemy2** - ORM with PostGIS spatial query support
+- **PostgreSQL/PostGIS** - Database with spatial extensions. Hosted on Amazon.
+- **AWS Lambda** - Serverless public access point. Documented in the [deployment](deployment/README.md) folder.
+
+### API - Design
+
+The central logic for API lives in [api.py](snowexsql/api.py)
+All measurement types must inherit from `BaseDataset`.
+
+### Database Schema
+
+Further explained in classes under [snowexsql/tables](snowexsql/tables)
+
+## Testing
+
+Need to pass with every code change.
+Command:
+
+```
+pytest tests/
+```
+
+# Skills
+Project specific skills are defined in [skills](skills/)
+
diff --git a/CLAUDE.md b/CLAUDE.md
new file mode 100644
index 0000000..43c994c
--- /dev/null
+++ b/CLAUDE.md
@@ -0,0 +1 @@
+@AGENTS.md
diff --git a/skills/db-query/SKILL.md b/skills/db-query/SKILL.md
new file mode 100644
index 0000000..e4aa157
--- /dev/null
+++ b/skills/db-query/SKILL.md
@@ -0,0 +1,30 @@
+---
+name: db-query
+description: Guide to answering SnowEx database query questions
+whenToUse: When the user asks how to query, access, or filter data from the SnowEx database
+---
+
+# SnowEx Query Skill
+
+## No setup required
+
+Use the preferred Lambda client. It works with zero configuration. No env vars, credentials, or AWS account.
+Do not tell users to set `SNOWEX_LAMBDA_URL` or any other variable.
+
+## Filtering
+
+Check `ALLOWED_QRY_KWARGS` of the `BaseDataset` class in `snowexsql/api.py` to get supported filtering criteria.
+Use `all_*` properties of query classes to inspect possible values to filter on.
+Allowed filter kwargs, available campaigns/sites/types change over time.
+
+## Query result
+
+All returned timestamps are in UTC.
+All returned coordinates are in WGS84 (EPSG:4326) lat/lon.
+
+### Limits
+
+The Lambda client returns a maximum of 1000 records per query.
+Advise users to use filters to narrow results rather than paginating.
+Use a count query to check how many records match a filter before running a full query and
+advise the user to narrow the filter if the count is too high.