Terminal, Git & AWS Command Cheat Sheet

Terminal

Everyday shell work: moving around, reading files, finding things, managing processes.

Navigate

• pwd

  Print the folder you're in.

• ls -la

  List all files, including hidden ones, with permissions, size and date.

• ls -lhS

  List files by size, largest first, in human units (K, M, G).

• cd <dir> · cd .. · cd - · cd ~

  Go into a folder · up one level · back to the previous folder · home.

• tree -L 2

  Show the folder structure two levels deep (brew install tree).

• open .

  macOS: open the current folder in Finder. On Linux use xdg-open.

• code .

  Open the current folder in VS Code.

Files and folders

• mkdir -p a/b/c

  Create nested folders in one go; no error if they already exist.

• touch file.txt

  Create an empty file, or update the timestamp of an existing one.

• cp -r src/ dest/

  Copy a folder and everything in it.

• mv old.txt new.txt

  Rename or move a file.

• rm -rf <dir>

  Destructive ·Delete a folder and everything in it, with no undo. Check the path first.

• ln -s <target> <link>

  Create a symbolic link (shortcut).

• chmod +x script.sh

  Make a script executable.

• chmod 600 key.pem

  Owner-only read/write access; required for SSH keys.

• chown user:group file

  Change a file's owner (usually needs sudo).

Read files

• cat file

  Print the whole file.

• less file

  Scroll through a large file. Keys: / search, n next match, G end, q quit.

• head -n 20 file · tail -n 20 file

  Show the first or last 20 lines.

• tail -f app.log

  Follow a log live as new lines are written.

• wc -l file

  Count lines. A quick row count for a CSV.

• diff -u a.txt b.txt

  Show line-by-line differences between two files.

Search

• grep -rn "text" .

  Search every file under this folder; show file and line number.

• grep -i -v "debug" app.log

  -i ignores case, -v keeps only the lines that DON'T match.

• rg "pattern"

  ripgrep: much faster grep that skips .gitignore'd files (brew install ripgrep).

• find . -name "*.csv" -mtime -1

  Find CSV files changed in the last day.

• find . -type f -size +100M

  Find files larger than 100 MB.

• fd pattern

  A simpler, faster find (brew install fd).

Text processing (pipes)

• cmd1 | cmd2

  Pipe: send the output of cmd1 into cmd2.

• cmd > out.txt · cmd >> out.txt

  Write output to a file (overwrite) · append to it.

• cmd 2>&1 | tee run.log

  Merge errors into output, show it on screen, and save it to a file.

• sort | uniq -c | sort -rn

  Count how often each line appears, most frequent first.

• cut -d, -f1,3 data.csv

  Take columns 1 and 3 from a comma-separated file.

• awk -F, '{s+=$3} END{print s}' data.csv

  Sum column 3 of a CSV.

• sed -i '' 's/old/new/g' file

  Replace text in place (macOS form; on Linux drop the '').

• xargs -n1 -P4 <cmd>

  Run a command on each input line, 4 at a time in parallel.

Processes and system

• ps aux | grep <name>

  Find a running process and its PID.

• top · htop

  Live CPU and memory by process (htop: brew install htop).

• kill <pid> · kill -9 <pid>

  Ask a process to stop · force-kill it.

• lsof -i :3000

  Show which process is using port 3000.

• kill $(lsof -t -i :3000)

  Destructive ·Free a stuck port by killing whatever is on it.

• cmd & · jobs · fg

  Run in the background · list background jobs · bring one back.

• nohup cmd > out.log 2>&1 &

  Keep a command running after you close the terminal.

• df -h · du -sh *

  Free disk space · size of each item in this folder.

• history | grep <word>

  Find a command you ran earlier. Ctrl+R searches interactively.

Environment

• echo $PATH | tr ':' '\n'

  Show each folder the shell searches for commands, one per line.

• which <cmd> · type <cmd>

  Show where a command lives, or whether it's an alias or function.

• export KEY=value

  Set an environment variable for this session.

• env | grep AWS

  Show environment variables matching a word.

• source ~/.zshrc

  Reload your shell config without opening a new terminal.

• alias gs='git status'

  Create a shortcut. Put it in ~/.zshrc to keep it.

Network and remote

• curl -sS https://api.example.com | jq

  Call an API and pretty-print the JSON.

• curl -X POST -H 'Content-Type: application/json' -d '{"a":1}' <url>

  Send a JSON POST request.

• curl -I <url>

  Show only the response headers.

• ping host · nc -zv host 5432

  Check a host is reachable · check a port is open.

• dig example.com +short

  Look up a domain's DNS records.

• ssh -i key.pem user@host

  Log in to a remote server with a key.

• scp file user@host:/path

  Copy a file to a server.

• rsync -avz --progress src/ user@host:/dest/

  Sync a folder, copying only what changed. Add --delete to mirror exactly.

• ssh -L 5433:db-host:5432 user@bastion

  Tunnel: reach a private database at localhost:5433 through a jump host.

Archives

• tar -czf out.tar.gz folder/

  Compress a folder.

• tar -xzf file.tar.gz

  Extract a .tar.gz file.

• zip -r out.zip folder/ · unzip file.zip

  Create or extract a zip file.

• gzip -k big.csv · gunzip big.csv.gz

  Compress a file (keeping the original) · decompress.

Sessions and scheduling

• tmux new -s work

  Start a named session that survives disconnects.

• tmux attach -t work

  Reattach to it. Ctrl+B then D detaches.

• crontab -e · crontab -l

  Edit or list scheduled jobs.

• 0 2 * * * /path/job.sh >> /tmp/job.log 2>&1

  Cron line: run a job every day at 02:00 and log its output.

Git

Version control: branching, committing, syncing, and undoing mistakes safely.

Setup

• git config --global user.name "Name"

  Set the author name on your commits (and user.email the same way).

• git config --global init.defaultBranch main

  Make new repos start on a branch called main.

• git clone <url>

  Download a repository.

• git init

  Turn the current folder into a new repository.

Daily loop

• git status -sb

  Short status: what changed and whether you're ahead or behind.

• git diff · git diff --staged

  Changes not yet staged · changes staged for the next commit.

• git add <file> · git add -p

  Stage a file · stage chosen chunks interactively.

• git commit -m "message"

  Commit what's staged.

• git commit --amend

  Fix the last commit's message or add forgotten files. Only before you push.

• git pull --rebase

  Get remote changes and replay your commits on top, keeping history straight.

• git push -u origin <branch>

  Push a new branch and set it to track the remote.

Branches

• git switch -c <branch>

  Create a branch and switch to it.

• git switch <branch>

  Switch to an existing branch.

• git branch -a

  List local and remote branches.

• git branch -d <branch>

  Destructive ·Delete a merged local branch. -D forces it.

• git fetch --prune

  Update remote branches and drop ones deleted on the server.

• git merge <branch>

  Merge a branch into the current one.

• git rebase origin/main

  Replay your branch on top of the latest main.

• git worktree add ../feat <branch>

  Check out a second branch in another folder at the same time.

History

• git log --oneline --graph --all -20

  A compact picture of recent history across branches.

• git log -p <file>

  Every change ever made to one file.

• git log -S "text"

  Find the commit that added or removed some text.

• git blame <file>

  Who last changed each line, and in which commit.

• git show <sha>

  Show one commit's changes.

• git reflog

  Everything HEAD pointed to recently. Use it to recover 'lost' commits.

Undo

• git restore <file>

  Destructive ·Discard unstaged changes to a file.

• git restore --staged <file>

  Unstage a file but keep the changes.

• git reset --soft HEAD~1

  Undo the last commit and keep its changes staged.

• git reset --hard HEAD~1

  Destructive ·Throw away the last commit AND its changes.

• git revert <sha>

  Add a new commit that undoes an old one. Safe on shared branches.

• git clean -fd

  Destructive ·Delete untracked files and folders. Preview first with -n.

• git stash · git stash pop

  Shelve uncommitted work · bring it back.

• git cherry-pick <sha>

  Copy one commit onto the current branch.

Remotes and releases

• git remote -v

  Show where this repo pushes and pulls.

• git tag v1.2.0 && git push --tags

  Tag a release and push the tag.

• git push --force-with-lease

  Destructive ·Force-push only if nobody else has pushed since. Never on main.

• gh pr create --fill

  GitHub CLI: open a pull request from this branch.

• gh pr checkout <number>

  Check out someone's pull request locally.

• gh run watch

  Watch the current CI run live.

AWS CLI

Add --profile <name> to target an account and --region <region> to target a region. Add --output table or a --query filter to make output readable.

Sign-in and identity

• aws configure

  Set up an access-key profile interactively.

• aws configure sso

  Set up an SSO profile.

• aws sso login --profile <name>

  Refresh an expired SSO login.

• aws sts get-caller-identity

  Check which account and user or role you're using. Always run this first.

• aws configure list-profiles

  List your profiles.

• export AWS_PROFILE=<name>

  Use one profile for the rest of this terminal session.

S3

• aws s3 ls · aws s3 ls s3://bucket/prefix/ --recursive --human-readable --summarize

  List buckets · list objects with sizes and a total.

• aws s3 cp file.csv s3://bucket/path/

  Upload a file. Swap the two arguments to download.

• aws s3 sync ./data s3://bucket/data/

  Destructive ·Upload only new or changed files. --delete removes extra files at the destination.

• aws s3 rm s3://bucket/path/ --recursive

  Destructive ·Delete everything under a prefix. Preview with --dryrun.

• aws s3 presign s3://bucket/file --expires-in 3600

  Make a download link that works for one hour.

• aws s3api head-object --bucket b --key k

  Show an object's size, type and metadata.

EC2 and servers

• aws ec2 describe-instances --query 'Reservations[].Instances[].[InstanceId,State.Name,Tags[?Key==`Name`].Value|[0]]' --output table

  List instances with their state and name.

• aws ec2 start-instances --instance-ids i-123

  Start an instance (stop-instances to stop it).

• aws ssm start-session --target i-123

  Open a shell on an instance with no SSH key or open port.

• aws ssm send-command --document-name AWS-RunShellScript --targets Key=instanceids,Values=i-123 --parameters commands='df -h'

  Run a command on an instance remotely.

Logs and monitoring

• aws logs tail /aws/lambda/<fn> --follow --since 10m

  Follow a log group live.

• aws logs filter-log-events --log-group-name <g> --filter-pattern ERROR

  Search a log group for a word.

• aws cloudwatch get-metric-statistics …

  Pull metric data points, such as CPU, for a time range.

IAM and secrets

• aws iam list-users · list-roles

  List users or roles.

• aws iam list-attached-role-policies --role-name <r>

  See which policies a role has.

• aws secretsmanager get-secret-value --secret-id <id> --query SecretString --output text

  Read a secret.

• aws ssm get-parameter --name /app/db_url --with-decryption

  Read a Parameter Store value.

Data services

• aws athena start-query-execution --query-string "SELECT …" --result-configuration OutputLocation=s3://bucket/athena/

  Run an Athena SQL query over S3 data.

• aws athena get-query-results --query-execution-id <id>

  Fetch that query's results.

• aws glue start-job-run --job-name <job>

  Start a Glue ETL job.

• aws glue get-job-runs --job-name <job> --max-items 5

  Check a Glue job's recent runs.

• aws glue get-tables --database-name <db>

  List tables in the Glue Data Catalog.

• aws redshift-data execute-statement --workgroup-name <wg> --database dev --sql "SELECT 1"

  Run SQL on Redshift Serverless with no database driver.

• aws emr list-clusters --active

  List running EMR (Spark) clusters.

• aws kinesis list-streams

  List Kinesis data streams.

• aws rds describe-db-instances --query 'DBInstances[].[DBInstanceIdentifier,DBInstanceStatus,Endpoint.Address]' --output table

  List databases with their status and endpoint.

Lambda and Bedrock

• aws lambda list-functions --query 'Functions[].FunctionName'

  List Lambda functions.

• aws lambda invoke --function-name <fn> --payload '{}' --cli-binary-format raw-in-base64-out out.json

  Run a function and save its response.

• aws bedrock list-foundation-models --region us-east-1 --query 'modelSummaries[].modelId'

  List the models Bedrock offers in a region.

• aws bedrock-runtime converse --model-id <id> --messages '[{"role":"user","content":[{"text":"hi"}]}]'

  Send one test prompt to a model (billed).

Cost and housekeeping

• aws ce get-cost-and-usage --time-period Start=<yyyy-mm-01>,End=<yyyy-mm-dd> --granularity MONTHLY --metrics UnblendedCost

  Spend for a date range, e.g. this month so far.

• aws cloudformation describe-stacks --query 'Stacks[].[StackName,StackStatus]' --output table

  List CloudFormation stacks and their status.

Data engineering

Databases, file formats, Python environments, containers, orchestration and big-data jobs.

PostgreSQL (psql)

• psql "postgresql://user:pass@host:5432/db"

  Connect to a database.

• \l · \dt · \d table · \dn

  Inside psql: list databases · tables · a table's columns · schemas.

• \x auto · \timing on

  Readable output for wide rows · show how long each query takes.

• \copy t TO 'out.csv' CSV HEADER

  Export a table (or a query in parentheses) to a local CSV.

• \copy t FROM 'in.csv' CSV HEADER

  Load a local CSV into a table.

• EXPLAIN (ANALYZE, BUFFERS) SELECT …;

  Show the query plan and real timings. The first tool for a slow query.

• pg_dump -Fc db > db.dump · pg_restore -d db db.dump

  Back up · restore a database.

DuckDB (fast local SQL on files)

• duckdb

  Open an in-memory analytics shell (brew install duckdb).

• SELECT * FROM 'data/*.parquet' LIMIT 10;

  Query Parquet or CSV files directly; no loading step.

• COPY (SELECT …) TO 'out.parquet' (FORMAT parquet);

  Convert or export to Parquet.

• DESCRIBE SELECT * FROM 'file.csv';

  See the column types DuckDB infers from a file.

JSON and CSV

• jq '.items[] | {id, name}' data.json

  Pick fields out of JSON.

• jq -r '.[] | [.id,.name] | @csv' data.json

  Turn a JSON array into CSV.

• csvlook data.csv | less -S

  View a CSV as an aligned table (pip install csvkit).

• csvstat data.csv

  Per-column stats: types, nulls, min and max.

• column -s, -t < data.csv | less -S

  Quick table view with no install.

Python environments

• python3 -m venv .venv && source .venv/bin/activate

  Create and turn on a project virtual environment.

• pip install -r requirements.txt

  Install a project's dependencies.

• pip freeze > requirements.txt

  Save exact installed versions.

• uv venv · uv pip install pandas

  uv: a much faster venv and pip replacement.

• python -m http.server 8000

  Serve the current folder over HTTP.

• jupyter lab

  Start notebooks in the browser.

Docker

• docker ps -a

  List containers, including stopped ones.

• docker compose up -d · docker compose down

  Start a stack in the background · stop and remove it.

• docker logs -f <container>

  Follow a container's logs.

• docker exec -it <container> bash

  Open a shell inside a running container.

• docker run --rm -p 5432:5432 -e POSTGRES_PASSWORD=pw postgres:16

  Throwaway local Postgres.

• docker build -t app:dev .

  Build an image from the Dockerfile here.

• docker system prune -a

  Destructive ·Free disk space: removes unused images, containers and cache.

Kubernetes

• kubectl get pods -n <ns>

  List pods in a namespace.

• kubectl logs -f <pod>

  Follow a pod's logs.

• kubectl describe pod <pod>

  See why a pod is failing: events and restarts.

• kubectl exec -it <pod> -- sh

  Shell into a pod.

• kubectl port-forward svc/<svc> 8080:80

  Reach a cluster service on localhost:8080.

Spark, Airflow and dbt

• spark-submit --master yarn --deploy-mode cluster job.py

  Submit a PySpark job to a cluster.

• pyspark

  Interactive Spark shell in Python.

• airflow dags list · airflow dags trigger <dag_id>

  List DAGs · start a run.

• airflow tasks test <dag> <task> <yyyy-mm-dd>

  Run one task locally without the scheduler.

• dbt debug

  Check dbt's connection and project setup.

• dbt run --select model_name+

  Build a model and everything downstream of it.

• dbt test · dbt build

  Run data tests · run models and tests together.

• dbt docs generate && dbt docs serve

  Browse lineage and docs in the browser.

Moving and checking data

• rsync -avz --partial src/ dest/

  Resumable folder copy.

• split -l 1000000 big.csv part_

  Split a huge CSV into 1M-line pieces.

• shasum -a 256 file

  Checksum a file to confirm a transfer arrived intact.

• iconv -f latin1 -t utf-8 in.csv > out.csv

  Fix text-encoding problems.

• file data.csv

  Detect a file's type and encoding.