The AWS terminal UI you hand your team instead of a page of AWS CLI commands.

Seamless Glance is a fast, read-only TUI for DevOps and cloud engineers. Point it at an AWS profile and it opens on a ranked list of findings (incidents, waste, and hygiene), with cost, account inventory, and service health one keystroke away. When you need to go deeper, it hands you the exact AWS CLI command or console link for the selected resource.

Website: seamlessglance.com

Every team that runs on AWS has the same wiki page: a list of aws ec2 describe-instances --query ..., aws cloudwatch describe-alarms --state-value ALARM, and aws elbv2 describe-target-health ... snippets that someone pastes into a terminal when something looks wrong. It works, but only for the person who already knows which command to run and how to read the JSON.

Seamless Glance replaces that page with one command:

glance

• Findings first. Alarms in ALARM, target groups with no healthy targets, security groups open to the world, idle EC2, stale Lambdas, untagged ECR images, and more, ranked by severity with a suggested next step.
• Cost in context. Month-to-date spend, forecast, a six-month trend, and a savings view that ties spend to the resources behind it.
• Pivot, do not paste. c shows (and can run) the AWS CLI command for the selected resource. o opens it in the AWS console. d describes it in place.
• Safe to hand out. It only makes read calls (Describe, List, Get), uses the standard AWS credential chain, and has no telemetry. A teammate can run it with a read-only role without being able to change anything.
• Built for the terminal. A single Rust binary, keyboard-driven, with region and profile switching in-app.

Press c: the exact AWS CLI command for the selected resource	Cost overview: spend, forecast, and the usage behind it
ECR: untagged images and missing lifecycle policies	Press d: describe any resource in place
CloudWatch alarms by state	Target group health
Security groups open to the world	Account inventory snapshot

_{Captured from v1.3.0 against a real account. Account IDs, resource IDs, and internal names are replaced with placeholders.}

brew install fells-code/seamless/seamless-glance

Detects your platform (Intel/ARM, glibc/musl) and installs the latest release:

curl -fsSL https://raw.githubusercontent.com/fells-code/seamless-glance/main/install.sh | bash

Pin a version with SEAMLESS_GLANCE_VERSION=1.2.3. The installer checks the binary against the release's SHA256SUMS.txt and refuses to install it if the checksum is missing or does not match. The Windows installer below does the same.

irm https://raw.githubusercontent.com/fells-code/seamless-glance/main/install.ps1 | iex

cargo install seamless-glance          # build from source
cargo binstall seamless-glance         # fetch the prebuilt release binary

Prebuilt binaries are published for macOS (arm64/x86_64), Linux (x86_64/aarch64, glibc and static musl), and Windows (x86_64). Both seamless-glance and the shorter glance alias are installed.

glance                       # uses your default AWS profile
glance --profile my-profile  # start on a specific profile

Glance resolves credentials exactly like the AWS CLI: environment variables, ~/.aws/config and ~/.aws/credentials, AWS SSO, and assumed roles. If aws sts get-caller-identity works in your shell, Glance will too.

Press ? at any time for help, / to jump to any view, and q to quit.

• Read-only. Glance only calls Describe, List, and Get style APIs. It never reads secret values (secretsmanager:GetSecretValue is not used).
• Local. It runs on your machine, talks directly to AWS API endpoints, and sends no telemetry or analytics anywhere.
• Fails closed. If a permission is missing, the affected view shows the access error and the rest of the app keeps working.

docs/iam-policy.json is a least-privilege policy that covers every call Glance makes. Attach it to a dedicated role for teammates who only need visibility. The ce: and pricing: actions are only needed for the cost views.

To report a vulnerability, follow the security policy rather than opening a public issue.

EC2, Lambda, and RDS also support a global region slot that aggregates across every enabled region. Other views are region-scoped.

Incident risk:

• CloudWatch alarms in ALARM
• target groups with zero healthy or some unhealthy targets
• load balancers with no active target path or zero healthy targets
• SQS queues with high visible or in-flight message counts
• RDS instances that are not available

Waste:

• running EC2 instances averaging below 5 percent CPU over 7 days
• stopped EC2 instances
• Lambda functions with high memory or no deploy in 180+ days
• ECR repositories with 20+ untagged images, or no push in 180 days and no pull in 90
• target groups with no load balancer and no registered targets
• API Gateway APIs with generic names or older than one year

Hygiene:

• security groups open to the world, or exposing sensitive ports publicly
• stopped EC2 instances with public IPs or production-like names
• EC2 instances missing Name, Owner, or Environment tags, and other resources missing an Owner tag
• production-like single-AZ RDS instances
• secrets without rotation, or with stale rotation
• SQS queues without a DLQ
• default VPCs still present
• deployed services with no matching CloudWatch alarms
• ECR repositories with images but no lifecycle policy

Enter on a finding opens the related service view. The findings roadmap tracks what is coming next, and adding a finding is a good first contribution.

Navigation:

Resource actions:

Actions always target the selected resource's own region, including in the global slot.

Glance captures the mouse so the wheel can scroll. To select text while it is running, hold a modifier while dragging (Shift in most Linux terminals, Option in iTerm2, Fn in macOS Terminal).

Press / and type a view name or alias:

• views: findings, account, cost, savings, ec2, ecs, ecr, rds, lambda, apigw, sqs, vpc, cw, sm, lb, tg, sg
• aliases such as overview, billing, api, queues, registry, secrets, and alarms
• region <name> (or rg <name>) to change region
• profile <name> (or pf <name>) to switch profile, or profile to open the picker
• theme <name> with autumn, winter, summer, spring, or developer

Press p (or run profile in the palette) to pick from the profiles in ~/.aws/config and ~/.aws/credentials. Selecting one rebuilds the AWS clients, keeps the current region, refreshes the active view, and remembers the choice for the next launch. What a profile can see is governed entirely by its IAM permissions, not by the app.

Issues and pull requests are welcome. Good places to start:

• Add a finding
• Add an AWS service view
• issues labeled good first issue

Build and check locally:

cargo run
cargo fmt --all
cargo clippy --all-targets -- -D warnings
cargo test --all

User-facing changes ship with a changeset:

npm install          # one-time, installs @changesets/cli
npm run changeset    # pick patch/minor/major and describe the change

On merge to main, a "version packages" PR is opened automatically. Merging it builds binaries for every platform, publishes the GitHub Release, updates the Homebrew tap, and publishes to crates.io. See the release process.

More reading:

• Architecture
• Development guide
• Findings roadmap
• AWS service roadmap
• Navigation strategy
• Contributing guidelines and code of conduct
• Agent guidance

Seamless Glance is free and open-source software, licensed under the Apache License, Version 2.0. You are free to use, study, share, modify, and build on it, including in commercial and closed-source work, under the terms of that license.

Copyright (C) 2026 Fells Code, LLC.