| Version |
Release date |
Release notes |
| 3.14.0 (latest) |
September 10, 2026 |
Added the --tui option, which runs the CLI
in a new terminal UI. When a shell command started with
background=true finishes, its result is now delivered to
the agent automatically on its next step. Completions are picked up
without the agent having to check on them. Fixed an issue where Agent Minutes double-counted a
turn's partial usage when a sub-agent was
billed. Fixed an issue where a turn whose stream closed without
completing hung or reported success. It now fails with an
error. Fixed atx ct per-repository analysis
progress reporting. A scoped container no longer overwrites the
progress of the repositories running alongside
it. Fixed atx ct remote Amazon EC2 jobs to
report why an SSM launch failed, instead of surfacing only the
probe's exit code.
|
| 3.13.0 |
September 2, 2026 |
Added atx ct remediation run as an alias for
atx ct remediation create, for consistency with
atx ct analysis run. atx ct remote network discover now prints a
next-step hint to run atx ct remote network create when it
finds no private subnets.
atx ct remote status now reports jobs waiting
for account concurrency as queued instead of
pending. This distinguishes a job that is waiting for
capacity from one whose state could not be read. Applies to both
Amazon EC2 and AWS Batch.
atx ct remote status now reports where a
failed job's container log is. For Amazon EC2 this is the log group and
stream, and for AWS Batch this is the CloudWatch stream. Previously
it showed only an exit code.
Changed resuming with --conversation-id to
apply -g, --configuration on top of the
session's stored configuration one setting at a time, instead of
replacing all of it. Settings you do not pass again are kept rather
than dropped. Fixed an issue where cancelling a remote analysis
terminated the container running it. The job now completes cleanly
instead of being reported as failed and retried by AWS
Batch. Fixed an issue where cancelling a remote analysis left its
containers polling until the wait timeout. Containers now end promptly,
and a cancelled job is no longer reported to AWS Batch as a failure
and retried. Fixed an issue where cancelling a security
analysis stopped only local polling while the code review jobs ran to
completion. The jobs are now stopped
service-side. Fixed an issue where atx ct analysis cancel
marked a running rapid-techdebt-analysis as cancelled
while the scan continued over the remaining repositories. Applies to
both local and remote runs. Fixed an issue where atx ct analysis cancel
marked a running agentic-readiness or modernization-readiness analysis
as cancelled while every remaining repository was still assessed.
Assessments that already finished keep their results. Applies to both
local and remote runs. Fixed atx ct remote status --group to report a
finished Amazon EC2 job as complete, so
--wait returns instead of polling to its
timeout. Fixed atx ct remote status to still report
what it observed when the Amazon EC2 instance is gone, instead of
failing outright. Fixed atx ct remote status to include waiting
jobs in its summary line, so the counts add up to the number of jobs
submitted. Fixed atx ct remote status --batch to show why
a job failed to start, instead of leaving it as
pending. Fixed atx ct remote status --batch to no
longer report not found for a batch that was just
submitted and is waiting for capacity. Fixed atx ct analysis get and
atx ct remediation status to settle a run whose
repositories have all finished. Fixed atx ct remote analysis
--resume-incomplete to no longer resubmit a repository that is
still queued, which could run the same repository
twice. Fixed analysis and remediation for self-hosted GitLab and
Bitbucket Data Center instances that are served under a URL subpath,
for example https://host/gitlab, or over plain HTTP.
Previously repository discovery succeeded but every clone and push
failed.
|
| 3.12.0 |
August 26, 2026 |
The CLI now supports atx ct commands on
Windows. Changed atx ct remote Amazon EC2 analyses to
default to 5 parallel workers. Added a -q, --quiet flag to
atx exec and atx custom def exec. The flag
mutes tool-call output for auto-approved tools, while tools that
require approval still render in full. Combine it with
--trust-all-tools to silence all tool
activity. Fixed file:// configuration paths on
Windows, including paths with spaces, so they now resolve
correctly. Fixed reading of files that have a UTF-8 byte-order mark,
so first-line edits and searches no longer fail. Fixed file overwrites to preserve the original line
endings, avoiding spurious whole-file changes on
Windows. Fixed analysis report parsing and patch application on
Windows to handle CRLF line endings. Fixed status symbols to fall back to plain text on older
Windows terminals that cannot display them. Fixed an issue where a kill-family command issued by the
agent through shell expansion could terminate the CLI's own process.
For example, commands such as kill $PPID or
kill $$ could affect the CLI. The shell
tool now checks these against the CLI's own process and its ancestor
chain before running them. Fixed clone operations to retry on transient network
failures. Fixed custom analyses to retry repository resolution when
it first returns no repositories. Fixed comprehensive and custom analyses to keep the local
clone needed for artifact upload. The CLI now rejects invalid --severity
filters on remote analyses with a clear error. Previously, it ignored
them. Fixed an issue where a completed recurring scheduled
analysis reported 0 findings even when findings existed.
atx ct analysis get and
atx ct analysis run --wait now report the correct
finding count. Fixed an issue where AWS Batch remote analyses could
fail with InsufficientFreeAddressesInSubnet when a subnet was
low on IP addresses. The dispatcher now waits for available IP
capacity before submitting each job. Fixed a dispatcher stall where AWS Batch and Amazon EC2
jobs could get stuck pending under high deferral volume. This could
happen with IP-constrained subnets or busy Amazon EC2 workers. Fixed remote provisioning to advertise the correct
region-qualified dispatcher IAM role name. This ensures the PassRole
grant an administrator sets up matches the role the dispatcher
uses. To apply the dispatcher fixes to an already-deployed
stack, run the following command:
atx ct remote update --mode ec2|batch --stack-name
<stack-name> [--execute] [--ack]
|
| 3.11.0 |
August 19, 2026 |
Added --repos to
atx ct remediation create for multi-repo remediation in
a single command. Added --transformation-name to
atx ct remote remediation to run a remediation directly
from a transformation definition without requiring existing
findings. Added --json output to
atx ct analysis run for machine-readable
results. Added an Agent Minutes cost summary to
atx ct analysis get and
atx ct remediation status. Added a local file export for
atx ct remote provision in dry-run mode: the generated
AWS CloudFormation templates are now written to a local file
(override the location with --out-dir) for review
before deployment. Fixed an issue where kill-family commands issued by the
agent could terminate the CLI's own process. Affected commands
include pkill -f node, xargs kill, and
find -exec kill. The shell tool now checks
these against the CLI's own process and its ancestor chain before
running them. Fixed an issue where GitLab source discovery used
repository topics as the language field. This could cause
repositories with special characters in their topics to fail
registration. Language is no longer inferred from
topics. Fixed an issue where atx ct commands that
failed could exit silently with empty output. The CLI now surfaces
actionable error messages. Fixed an issue where finalizing a local analysis run
could error when the service reported the run was already
terminal. Fixed an issue where
atx ct remote update did not update the shared
dispatcher infrastructure. As a result, upgrading the CLI and running
remote update continued to use the previously deployed
dispatcher. remote update now reconciles the
dispatcher. Fixed an issue where remote analysis runs (Amazon EC2 or
AWS Batch) could report a false failure. This occurred when all
repositories in a scoped run had already completed on the
service. Fixed an issue where atx ct repository
discovery could silently drop repositories the service rejected. The
CLI now counts and reports skipped repositories ("N repositories
were skipped"), with an overflow summary for large
organizations. Changed user-facing error and warning messages to refer
to AWS Transform and state a clear next step (refresh credentials,
retry, or contact support). Messages no longer reference internal
service names.
|
| 3.10.0 |
August 10, 2026 |
Added support for
background=true in the shell tool for
long-running or non-terminating commands (dev servers, build
watchers, tail -f). It returns immediately with a
process ID instead of waiting for the command to exit. Up to 5
background processes can be tracked at once. Added new check_background_shell and
stop_background_shell tools to monitor and terminate
background processes. Added llmConfig to the CLI configuration
allowlist, enabling users to customize LLM settings via config
files. Added --mode aws-managed to
atx ct schedule create and
atx ct remote analysis to run analyses using fully
managed AWS infrastructure without provisioning your own
Amazon EC2 or AWS Batch resources. Added --schedule-id and
--display-name options to
atx ct analysis to filter results by schedule and
identify analyses by a human-readable label. Added --region support for
--mode aws-managed operations, allowing you to target
a specific AWS Region for managed analysis runs. Improved reliability of
atx ct schedule get, enable,
disable, and delete
commands. Improved accuracy and consistency of
atx ct schedule list results. Added --stack-name to
atx ct schedule create to pin a schedule to a
specific target stack. Added new atx ct analysis list-artifacts
and atx ct analysis get-artifact commands for
retrieving output artifacts from AWS-managed infrastructure
analysis runs. Fixed an issue where background processes tracked by
the CLI were not properly terminated on CLI exit, which could leave
orphaned processes running. Fixed an issue where
atx ct schedule create did not reject the
--workers and --batch-name flags when
used with --mode aws-managed. Fixed inconsistent wording in schedule not-found error
messages. Fixed an issue where endpoint overrides were not
resolved consistently across all CLI commands. Improved accuracy of duplicate schedule name
detection. Fixed an issue where
atx ct remote analysis --resume did not resume a
partially-completed batch run. The flag is now
--resume-incomplete, which re-runs only the
repositories that did not finish (failed, interrupted, or not yet
started) and skips completed ones. Updated atx ct schedule --help to
include AWS Managed Analysis run
offerings.
|
| 3.9.0 |
July 31, 2026 |
Improved the performance of
atx custom def export-ki-markdown on transformation
definitions that contain many knowledge items. The command no
longer makes a redundant service call for each knowledge item, so
it runs significantly faster and no longer times out on large
definitions. Also fixed cases where an export could omit knowledge
items or report an inconsistent total. Fixed an issue where agentic-readiness and
modernization-readiness analyses run with
atx ct analysis run could report zero findings for
certain report formats. Findings are now recovered
reliably. Fixed an issue where scheduled analyses created with
atx ct schedule create against GitLab, Bitbucket, or
local sources could fail on every run while trying to retrieve
credentials. The source provider is now determined automatically
from your registered source, and the --provider option
is no longer required. Fixed an issue where security analyses run in remote
(AWS Batch or EC2) environments could incorrectly report that the
Security Agent was not configured. The CLI now detects a configured
Security Agent correctly and reports transient credential or
throttling errors more gracefully.
|
| 3.8.0 |
July 29, 2026 |
Added the --recurrence option to
atx ct schedule create, so you can set a schedule's
cadence in plain terms (daily,
weekly:<MONDAY..SUNDAY>, or
monthly:<1..28>) instead of writing a raw cron
expression. Added the --existing-scheduler-role-arn
option to atx ct remote provision, so you can reuse
an IAM role that your administrator has already created for the
scheduler instead of having the CLI create one during
provisioning. Added the
--existing-instance-profile-name option to
atx ct remote provision --mode ec2, so you can attach
an IAM instance profile that your administrator has already created
instead of having the CLI create the transform role and instance
profile. Improved transformation selection guidance so that it
now uses detection metadata, such as severity and matching
criteria, when available. This helps surface the transformations
that best fit your project and their relative
priority. Fixed an issue where canceling a scheduled remote
analysis stopped the running containers but could leave the
analysis stuck in a running state. Canceling now
reliably marks the analysis as canceled. Fixed an issue where atx ct schedule delete
could leave a lingering analysis record behind. Deleting a schedule
now also removes the underlying scheduled
analysis.
|
| 3.7.0 |
July 22, 2026 |
Added the atx ct remote command group for
running analyses and remediations remotely on EC2 or AWS Batch. You
can provision, update, and tear down the remote execution
infrastructure, check submission status, detect deployed stacks, and
manage source credentials in AWS Secrets Manager. Added the atx ct schedule command group for
recurring scheduled analysis and remediation jobs. You can create,
list, enable, disable, and delete schedules (job options mirror
remote analysis and remote remediation),
plus setup and teardown for the scheduler
infrastructure. Schedules capture local source bundles at create
time. Improved atx ct analysis run to consume
tech-debt-quick matching rules from the transformation package
registry. Improved long-running atx ct commands to
show progress output, including per-repo progress during analysis
runs (including security analyses) and per-repo failure reasons in
remediation status. Fixed several remote execution reliability issues. EC2
status now shows findings and pull request results, matching Batch;
EC2 updates no longer fail on stacks missing new parameters; and
additional fixes address exit codes, transform handling, tag-based
detection, and JSON output.
|
| 3.6.0 |
July 17, 2026 |
Added the atx custom def learnings
subcommand, which opens an interactive TUI session to view and
manage the learnings automatically generated for a transformation
definition based on previous runs. Added usage telemetry collection during transformation
execution. To disable telemetry, set the
ATX_DISABLE_TELEMETRY=true environment variable. For
more information, see Usage Telemetry. Fixed an issue on Windows where file operations
using ~ (home directory) paths resolved to an
unexpected directory. This caused file read/write failures
during a session. Fixed an issue where MCP server and tool names
containing special characters could cause connection failures.
The CLI now automatically sanitizes names to use only valid
characters.
|
| 3.5.0 |
July 8, 2026 |
Added native Windows support. You can now run the
AWS Transform Custom commands directly on Windows. (The atx ct commands
are not yet supported on Windows) Added steering inputs to atx ct analysis
run so agentic-readiness and modernization-readiness
analyses reflect your preferences. Use --context
for strategic framing, --agent-scope to set agent
access for agentic-readiness, and
--prefer/--avoid to steer
modernization recommendations toward or away from specific
technologies. Transformation definitions in the skill format can
now include a scripts/ folder, letting you bundle
helper scripts alongside your SKILL.md and
references. Added --wait to atx ct
remediation create, which waits until the remediation
completes. Added headless mode for one-shot execution. Run
atx -x "prompt" -t to pass a prompt, get a
response, and exit with no interactive prompts. Headless mode
requires trusting all tools.
|
| 3.4.1 |
July 2, 2026 |
|
| 3.4.0 (not recommended) |
June 29, 2026 |
In interactive sessions, the CLI might stay on the
"Thinking" prompt (shown with an ellipsis in the UI)
after the agent finishes responding. This prevents you from
entering your next message. To resolve this issue, upgrade to
version 3.4.1 or later by running
atx update.
Added atx ct schema, a
machine-readable manifest of the full atx ct
command surface for scripting and automation. Added token scope validation when you add or
update a source, surfacing credential or permission issues
immediately instead of failing later during
analysis. Improved auto-remediation pull requests to include
finding context and support draft mode for more reviewable
PRs. Improved CLI progress feedback with real-time
context about the current activity, estimated agent minutes,
and sub-agent activity. Added configurable checkpoint commit author
identity through the ATX_GIT_COMMITTER_NAME and
ATX_GIT_COMMITTER_EMAIL environment
variables. Various small bug fixes and
improvements.
|
| 3.3.0 |
June 18, 2026 |
Added consistent --json and
--help support across atx ct
commands. Fixed security agent setup so the agent space is
created during setup and removed on teardown. Various small bug fixes and
improvements.
|
| 3.2.0 |
June 15, 2026 |
Added continuous modernization commands
(atx ct) to the standard CLI distribution for
all users. Improved sub-agent execution to show a
"Thinking..." spinner while waiting for server
responses, providing clear visual feedback that work is in
progress. Improved spinner to automatically restart with
"Working on it..." after 3 seconds of silence
following agent output, so the terminal never appears frozen
between processing steps. Improved thinking spinner with a subtle shimmer
animation for a more polished visual
appearance.
|
| 3.1.0 |
June 3, 2026 |
Added ATX_MAX_VALIDATION_LOOPS
environment variable to configure maximum validation
loops. Fixed update check no longer shows false
"update available" warnings when the installed
version is already newer than the registry
version. Fixed delete
--transformation-name "" no longer shows a
nonsensical confirmation prompt; empty values are now rejected
with a clear validation error. Fixed unrecognized subcommands now show
"unknown command" instead of a confusing internal
error message. Fixed version comparison to use numeric comparison
instead of string inequality for update
checks.
|
| 3.0.0 |
May 21, 2026 |
Added pre-flight transformation name validation
that suggests similar names from your registry when a name is
not found, helping catch typos without needing a separate list
command. Added optional explanations for shell commands,
showing what they do and why. Improved trust settings banner to include more
information about tool permissions. Added learning-disabled banner when using the
do-not-learn flag, giving visibility to the learning feature
and how to enable it. Breaking change:
Node.js 22 or later is now required. Upgrade from
https://nodejs.org/ if you are using an older
version.
|
| 2.0.0 |
May 14, 2026 |
Added remote MCP server support — connect to
HTTP and HTTPS MCP servers with automatic SSE fallback, Bearer
token authentication with environment variable expansion in
headers, and proxy support
(HTTP_PROXY/HTTPS_PROXY). Added atx resume subcommand as alias
for atx --resume. Added grep tool for multi-file text
search with output control — supports regex patterns, file
glob filtering, case-insensitive search, configurable
depth/result limits, and output truncation. Improved MCP implementation to use the standard MCP
SDK for more consistent behavior and easier support for future
MCP features. Improved shell output handling — output is now
gracefully truncated at 30K characters per stream instead of
causing a hard error, preserving the beginning and end of output
for context. Fixed an issue where Ctrl+C during an interactive
session could prematurely exit the CLI on Linux/macOS instead of
aborting the current operation and returning to the
prompt. Fixed trust rejection message to show a dedicated
reason for why trust is unavailable, separate from why the
command needed authorization. Fixed dangerous command warning no longer mentions
trusting when the command is in
alwaysPromptCommands.
|
| 1.10.0 |
April 29, 2026 |
Improved tool descriptions with cross-tool guidance,
helping the agent choose the best tool for each task and reducing
unnecessary shell usage. Added alwaysPromptCommands in
trust-settings.yaml — configure shell command
patterns that always require explicit permission, regardless of
trust settings. Supports glob wildcards and matches against each
sub-command in compound expressions.
trustedShellCommands can override these patterns.
Not enforced in non-interactive mode to avoid breaking automated
pipelines. Fixed relative paths not being resolved into absolute
paths.
|
| 1.9.0 |
April 23, 2026 |
Improved tool approval prompts — each option
now shows what it does inline, a “Why” message
explains why approval is needed, and trust confirmation includes
a hint about trust-settings.yaml for permanent
trust.
|
| 1.8.0 |
April 16, 2026 |
Changed agent output prefix from > to
ATX: for clearer visual distinction between user
input and agent output. Fixed display of resume command when CLI retries are
exhausted. Fixed resume from a different working directory. In
interactive mode, prompts for corrected path; in non-interactive
mode, fails fast with a descriptive error.
|
| 1.7.1 |
April 9, 2026 |
Fixed SSO login failure ("Session token not found or
invalid") when the credential provider region differs from the
profile's sso_region. Fixed credential caching so long-running sessions no
longer fail when credentials are rotated on
disk.
|
| 1.7.0 |
April 2, 2026 |
Fixed exec command to exit with code 1
on input validation errors, making the CLI reliable in scripts
and CI/CD pipelines. Added display of warning messages from the
service. Added MFA prompt support for AWS credential
profiles. Fixed STS credential calls to respect HTTP(S) proxy
settings.
|
| 1.6.0 |
March 18, 2026 |
Added --limit <minutes> option to
set an Agent Minutes budget for exec and
interactive commands. When the limit is reached,
the CLI displays usage and exits with resume
instructions. Fixed logging of the initial user message in
conversation logs.
|
| 1.5.0 |
March 11, 2026 |
Added support for concurrent task execution, enabling
faster transformations. Added Agent Minutes usage display at the end of each
conversation, and /usage command in interactive
mode to check usage at any time.
|
| 1.4.0 |
March 2, 2026 |
Improved error messages when saving transformation
definitions with unsupported files. Added timeout to service calls to prevent the CLI
from hanging when the connection is
interrupted. Fixed CLI process hang after successful
completion. Fixed region validation when resuming
conversations.
|
| 1.3.0 |
February 19, 2026 |
|
| 1.2.0 |
February 9, 2026 |
Added support for proxy configuration via environment
variables (https_proxy, no_proxy,
etc.). General stability improvements and bug
fixes.
|
| 1.1.1 |
January 16, 2026 |
Bug fixes and stability improvements.
|
| 1.1.0 |
January 13, 2026 |
Added AWS CLI region configuration support, early credential
validation checks, and fixed campaign resume
functionality.
|
| 1.0.1 |
December 18, 2025 |
Graceful shutdowns, cleaner error logs, and general
improvements.
|
| 1.0.0 |
December 1, 2025 |
Initial release of the AWS Transform CLI.
|
