awsmux

awsmux

Run one AWS CLI command across your whole fleet in parallel: hundreds of accounts in seconds, safely.

One command fanned out across every account and region at once (100 parallel workers by default, --concurrency raises it), identities verified before anything runs, results merged into one stream. No more shell loops that take a coffee break to crawl the fleet. And anything that mutates is stopped by an approval boundary that even an AI agent with your admin credentials cannot talk its way past. The demo above is a real terminal against the bundled 100-account sandbox fleet; replay it yourself in one command with make build fleet-up.

And it is measurably cheaper and faster for agents. In a 150-session, three-arm benchmark (identical Claude Opus 4.8 agents, identical prompts, same 100-account fleet), the awsmux arm was cheaper and faster than a raw-shell aws CLI arm in every cell: 1.3x to 2.9x cheaper, 2.3x to 5.4x faster, and up to 7.4x fewer output tokens, at a flat 4 turns whatever the fleet size (the CLI arm matched that only at 10 accounts, then grew to 10.5; cost and output-token differences are all Holm-adjusted p < 0.05). 138 of the 150 sessions passed environment validation, and all 12 exclusions were awsmux's own fault: an MCP startup race under concurrency, so they fell entirely in the MCP-bearing arms.

task	fleet	cost per run (cli vs awsmux)	wall time
enumerate all VPCs	10 accounts	$0.090 vs $0.068 (1.3x)	34s vs 13s
enumerate all VPCs	50 accounts	$0.239 vs $0.120 (2.0x)	84s vs 23s
enumerate all VPCs	100 accounts	$0.274 vs $0.193 (1.4x)	94s vs 41s
find the world-open group	50 accounts	$0.229 vs $0.079 (2.9x)	75s vs 14s
find the world-open group	100 accounts	$0.152 vs $0.098 (1.5x)	90s vs 17s

Full design, statistics, and caveats: docs/BENCHMARK.md.

Try it on 100 accounts (none of them real)

Prerequisites: Go, Docker, and the aws CLI.

make build fleet-up && source .tmp/fleet/env.sh

That builds ./bin/awsmux, boots a pinned LocalStack container, and provisions a fictional 100-account fleet (10 teams, prod and stage, 5 shards, 3 regions) to break for fun. The real aws CLI talks to a real emulated AWS on localhost, every profile is its own account (plus one planted admin-legacy duplicate for --dedupe to catch), and what you change actually persists. Zero credentials, zero real AWS, zero risk.

./bin/awsmux targets --profiles 'payments-*'     # verified identities, per account
./bin/awsmux run --dedupe --format jsonl -- ec2 describe-vpcs --query 'Vpcs[].VpcId'
                                                 # all 100 accounts in a few seconds
./bin/awsmux run --profiles '*-prod-*' -- ec2 describe-security-groups \
  --filters Name=ip-permission.cidr,Values=0.0.0.0/0   # find the world-open group
./bin/awsmux plan --profiles payments-prod-1 -- ec2 revoke-security-group-ingress \
  --group-name legacy-bastion --protocol tcp --port 22 --cidr 0.0.0.0/0

That last one is destructive, so it will not just run: you get an immutable plan to approve first. Apply it with the token and re-run the hunt; the finding disappears for real, because the sandbox is a real (emulated) AWS. Try editing the plan file between approve and apply and watch the hash check refuse it.

make e2e builds the binary, provisions the fleet, and smoke-tests these same beats end to end, from STS-verified discovery through the approval gate to a plan / approve / apply roundtrip and a tampered plan refused by its hash (full list in docs/ARCHITECTURE.md); make fleet-down removes the container.

Real fleet

Same commands, pointed at your own AWS setup instead of the sandbox environment. Zero configuration of its own: awsmux discovers profiles from your existing shared config file (~/.aws/config) and shared credentials file (~/.aws/credentials), honoring AWS_CONFIG_FILE and AWS_SHARED_CREDENTIALS_FILE. SSO, static keys, and credential_process profiles all work unchanged, because awsmux always executes through the aws CLI and verifies every identity with STS before running anything:

awsmux doctor                       # first run? verify aws CLI, files, profiles
awsmux targets --regions us-east-1,us-west-2

awsmux run --profiles 'prod-*' --exclude '*-sandbox' --format jsonl \
  -- ec2 describe-instances --query 'Reservations[].Instances[].InstanceId'

awsmux plan -- ssm put-parameter --name /app/flag --value on --type String
awsmux approve plan-01k...          # prints a one-time token, never stored
awsmux apply plan-01k... --approval-token <token>

Not seeing your profiles? awsmux doctor shows exactly which files were checked, how many profiles each contributed, and whether the aws CLI and state directory are usable. awsmux targets reports where each profile came from (a SOURCE column in table mode, a source field in jsonl: config, credentials, or both).

Useful flags on run, and on apply / replay too unless noted: --concurrency (default 100: fleet-wide fan-out is the point; each in-flight target is one aws CLI subprocess), --timeout 30s, --max-errors N, --stop-on-access-denied, --output-dir (one result file per target; not on replay), --interactive (checkbox target picker; run only). Target selection (--profiles, --exclude, --regions, --preflight, --dedupe, --ou, --account-tag, --org-role, --org-profile, --org-refresh) applies to run, plan, and targets; --dedupe collapses targets that resolve to the same account, principal, and region, and it runs the STS preflight to find them even under --preflight=false. --show-unreachable is also accepted on all three but only targets acts on it. Every run lands in awsmux history and can be re-run with awsmux replay, which re-selects its targets from the stored execution.

Selecting by AWS Organizations

--ou and --account-tag select targets by org structure instead of profile name. Both filter on the account id STS verified during preflight, never on a profile name, and both match nested OUs: --ou eng/prod also selects eng/prod/db.

awsmux targets --ou eng/prod
awsmux targets --account-tag env=prod --show-unreachable
awsmux run --ou eng/* -- ec2 describe-instances

awsmux enumerates the org through the AWS CLI and caches the tree for an hour (--org-refresh to bypass). When your everyday credentials cannot call the Organizations API, --org-role <arn> assumes a role for the enumeration only; it never changes how targets execute, which is always aws --profile <name>.

Org accounts matching the filter that no local profile reaches are reported as unreachable rather than hidden, so awsmux targets --ou eng/prod tells you both what you can reach and what you cannot. If enumeration fails, the command errors out: an org selector never degrades into an unfiltered fan-out.

Give it to your AI agent

Two commands, zero configuration:

go install github.com/0hardik1/awsmux@latest
claude mcp add --scope user awsmux -- awsmux mcp

There is nothing to configure. awsmux mcp serves the agent interface over stdio via the Model Context Protocol: no port, no daemon, no config file, no credentials of its own. It discovers the same profiles your terminal already has (shared config and credentials files; SSO, static keys, and credential_process all pass straight through) and verifies every identity with STS before anything runs. It even locates the aws CLI in well-known install locations when an MCP client spawns it with a bare GUI PATH. Check the connection with /mcp inside Claude Code, then ask the agent "which AWS accounts can you see?" (Drop --scope user to register for the current project only. No Go toolchain? git clone + make build gives you ./bin/awsmux.)

Claude Desktop: Settings > Developer > Edit Config, then add to claude_desktop_config.json (GUI apps do not read your shell PATH, so use the absolute path from which awsmux):

{
  "mcpServers": {
    "awsmux": { "command": "/absolute/path/to/awsmux", "args": ["mcp"] }
  }
}

Other MCP clients (Cursor, Windsurf, Zed, ...): same shape everywhere, a stdio server with command awsmux and args ["mcp"].

The agent gets five structured tools (list_aws_targets, plan_aws_operation, execute_aws_plan, get_aws_execution, cancel_aws_execution) instead of a raw shell, and it uses them on its own: once registered there is nothing to invoke, you just ask ("find every security group open to the world in prod") and the model reaches for the tools automatically. Read-only operations execute freely. Anything else comes back as a plan: the agent hands you a plan ID, you run awsmux approve <plan-id> in your own terminal, and you paste the one-time token back into the chat for the agent to apply. The token binds to the plan's sha256 hash, so the agent cannot alter an approved plan or execute it twice.

Want to watch an agent hit the boundary with zero blast radius? Register the sandbox fleet instead: after make build fleet-up, point the server at the fleet's config via environment variables (values printed in .tmp/fleet/env.sh):

claude mcp add awsmux-fleet \
  --env AWS_CONFIG_FILE=$PWD/.tmp/fleet/aws-config \
  --env AWS_SHARED_CREDENTIALS_FILE=$PWD/.tmp/fleet/aws-credentials \
  --env AWSMUX_HOME=$PWD/.tmp/fleet/home \
  -- $PWD/bin/awsmux mcp

The agent gets the 100-account LocalStack fleet over MCP; let it break anything it likes.

The cost and speed numbers in the table at the top come from exactly this setup. One more result from the same benchmark: when an agent was handed both a shell and the MCP tools, it picked awsmux on its own and kept nearly all of the margin.

Safety model

Class	Examples	Policy
read_only	describe-, list-, get-*, s3 ls	runs freely
mutating	create-, put-, update-*, s3 cp, s3 presign	--yes, interactive confirm, or approved plan
destructive	delete-, terminate-, revoke-*, s3 rm, s3 mv	never --yes; typed confirm or plan/approve/apply
unknown	anything unrecognized	treated as mutating

Where the verb convention lies, awsmux overrides it. sts calls that mint credentials (assume-role*, assume-root, get-session-token, get-federation-token), s3api operations that write a local outfile (get-object, get-object-torrent, select-object-content), and s3 presign (the URL it prints is a bearer credential for the object) are all mutating despite their read-style names, and s3 mv is destructive because it deletes the source. Where the risk lives in an argument rather than the name, the argument decides: s3 sync is mutating, but s3 sync --delete is destructive.

Stable exit codes for CI and agents: 0 all succeeded, 1 some failed, 2 selection/config error, 3 approval required or rejected, 4 stopped by threshold.

The 03:00 story

A scanner pages: SSH is open to the world somewhere in the fleet. Your agent sweeps 14 verified accounts in one read-only call, finds sg-0a1b2c3d in payments-prod, and proposes the revoke. The revoke is destructive, so the engine hands back a plan instead of running it. You read the plan on your phone: exact command, exact account, verified principal, hash. You approve, paste one token, and go back to sleep while the fix lands in history with the hash attached.

You can replay this exact story against a fleet of fakes right now: make build fleet-up.

How it compares

	awsmux	awsrun	aws-vault / granted / awsume
Fleet execution across accounts	yes	yes	no (credential tools)
STS identity preflight per target	yes	no	n/a
Risk classification + approval boundary	yes	prompt only	n/a
Agent interface (MCP)	yes	no	no
100-account LocalStack test fleet (make fleet-up)	yes	no	no
History + replay	yes	no	no
Runtime	single Go binary	Python + plugins	Go / varies

The credential tools are complementary: awsmux happily executes through profiles they manage.

Docs

• Architecture: the engine, the plan boundary, the executor, test-fleet internals.
• Benchmark: the three-arm agent-cost methodology and results behind the table above.

Development

make setup                 # git hooks + pre-warm the pinned linter
make build test lint       # or: go build ./... && go vet ./... && go test ./...
make fleet-up e2e          # LocalStack fleet + end-to-end smoke test
make fleet-down clean

CI runs go test and make build on ubuntu and macos, gofmt + go vet, golangci-lint repo-wide, the LocalStack e2e smoke test on ubuntu, and a Conventional Commits check on PR titles (the commit-msg hook enforces the same convention locally).

The demo GIF at the top is a real Ghostty session against the LocalStack fleet, recorded with Kap.

Dependencies: stdlib plus cobra. Roadmap: policy packs, Homebrew tap and release binaries.

Licensed under Apache-2.0.