286 Commits
Author SHA1 Message Date
utsav-oaiandutsav-oai a69515510f feat(cli)!: support incremental exploitability investigations (#2745)
* feat(cli): add independently reviewed dependency checks

* Compare frozen Python advisory versions as releases

* fix(cli): bind unnamed root manifests to observed services

* test(cli): evaluate dependency investigations across public fixtures

* test(sdk): bound the budget wait after the mock scan starts

* test(cli): exercise MCP evidence and missing caller access

* Bind graded citations to the approved use evidence

* test: restore dependency investigation quality controls

* fix(cli): restore dependency investigation and report compatibility

* test: grade malformed dependency baseline answers

* fix(cli): scope headless tool access to the investigator

* test(cli): lock down reviewer and external evidence tool access

* fix(cli): keep linked evidence tied to its originating call

* Reconcile application mappings and expose original evidence hashes

* test: verify selected catalog rows against original replies

* fix(cli): recover incomplete reviews and preserve selected evidence rows

* fix(cli): preserve wire evidence and verify finding citations

* Finalize from approved submissions and explain remaining uses

* fix(cli): restore coordinator recovery, independent review and report contracts

* Review withdrawals under the original dependency assessment ID

* test(cli): compare trace progress with completed search scopes

* Recheck dependency coverage after concurrent investigation changes

* Preserve exploratory scope explanations during dependency review

* feat(cli): track shared source work in dependency investigations

* Include shared source work in final coverage checks

* Require deciding evidence on the caller it establishes

* Preserve generic MCP search outcomes and scopes in dependency reports

* Share pinned source notes between dependency investigators

* Ship the source worklist and search audit modules

* feat(cli)!: support incremental exploitability investigations

BREAKING CHANGE: use next to read the full captured stack. The read-stacks command has been removed.

* refactor(cli): name the dependency coordinator for its ability to resume

* refactor(cli): simplify the tools for dependency investigations

* fix(cli): include incremental investigation modules in the package inventory

* fix(cli): keep investigation tool arguments visible to Codex and reuse the evidence store

* fix(cli): send missing verdict citations back to the reviewer

* fix(cli): verify caller bindings and input flow across investigation steps

* fix(cli): require direct citations for asserted runtime facts on every use

* test(cli): verify failed catalog lookups without a timed source scan

* test(cli): release the legacy report statement before Windows cleanup

---------

Co-authored-by: utsav-oai <348536218+utsav-oai@users.noreply.github.com>
2026-11-19 22:45:21 -05:00
utsav-oai bcf39188a3 fix(cli): assess captured stacks against public advisory conditions (#2760)
* fix(cli): assess captured stacks against public advisory conditions

Resolve advisory records once per captured-stack investigation and extract necessary conditions with supporting quotations. Share the combined conditions with the investigator and proof validator, and freeze the evidence for replay.

* refactor(cli): share advisory trigger guidance across investigations

* fix(package): include the stack advisory module in the distribution manifest

* fix(cli): review captured-stack investigations before publication

Verify cited checks for unresolved questions during assessment and finalization. Add independent stack review and share the repository workflow’s one-shot correction loop.

* fix(cli): expose each stack report in the command output schema

* refactor(cli): use one report format for captured stacks

* fix(evals): grade stack reports against their review and advisory evidence
2026-11-19 21:28:34 -05:00
utsav-oai ff1123120b feat(cli): render exploitability reports in Markdown and interactive HTML (#2762)
* feat(cli): render exploitability reports in Markdown and interactive HTML

* fix(package): include exploitability report modules in the distribution manifest

* fix(cli): use declared format choices and parse report markup in tests

* fix(cli): restore keyboard focus after leaving the report walkthrough

* test(package): size the bounded archive fixture from its contents

* feat(cli): show captured frames and matching source inline in Markdown

* feat(cli): explain how to supply context for unknown verdicts
2026-11-19 17:36:23 -08:00
dependabot[bot]andMichael D'Angelo 2d4442acc6 chore(deps): update Papa Parse and Prettier and check Node engine compatibility (#2754)
* chore(deps): bump the third-party group across 2 directories with 6 updates

Bumps the third-party group with 1 update in the /plugins/codex-security/mcp-app directory: [prettier](https://github.com/prettier/prettier).
Bumps the third-party group with 6 updates in the /sdk/typescript directory:

| Package | From | To |
| --- | --- | --- |
| [prettier](https://github.com/prettier/prettier) | `3.9.10` | `3.10.1` |
| [@inquirer/checkbox](https://github.com/SBoudrias/Inquirer.js) | `5.2.8` | `5.3.0` |
| [@inquirer/confirm](https://github.com/SBoudrias/Inquirer.js) | `6.3.5` | `6.4.0` |
| [@inquirer/input](https://github.com/SBoudrias/Inquirer.js) | `5.1.9` | `5.2.0` |
| [@inquirer/search](https://github.com/SBoudrias/Inquirer.js) | `4.3.6` | `4.4.0` |
| [papaparse](https://github.com/mholt/PapaParse) | `5.7.0` | `5.7.1` |



Updates `prettier` from 3.9.10 to 3.10.1
- [Release notes](https://github.com/prettier/prettier/releases)
- [Changelog](https://github.com/prettier/prettier/blob/main/CHANGELOG.md)
- [Commits](https://github.com/prettier/prettier/compare/3.9.10...3.10.1)

Updates `prettier` from 3.9.10 to 3.10.1
- [Release notes](https://github.com/prettier/prettier/releases)
- [Changelog](https://github.com/prettier/prettier/blob/main/CHANGELOG.md)
- [Commits](https://github.com/prettier/prettier/compare/3.9.10...3.10.1)

Updates `@inquirer/checkbox` from 5.2.8 to 5.3.0
- [Release notes](https://github.com/SBoudrias/Inquirer.js/releases)
- [Commits](https://github.com/SBoudrias/Inquirer.js/compare/@inquirer/checkbox@5.2.8...@inquirer/checkbox@5.3.0)

Updates `@inquirer/confirm` from 6.3.5 to 6.4.0
- [Release notes](https://github.com/SBoudrias/Inquirer.js/releases)
- [Commits](https://github.com/SBoudrias/Inquirer.js/compare/@inquirer/confirm@6.3.5...@inquirer/confirm@6.4.0)

Updates `@inquirer/input` from 5.1.9 to 5.2.0
- [Release notes](https://github.com/SBoudrias/Inquirer.js/releases)
- [Commits](https://github.com/SBoudrias/Inquirer.js/compare/@inquirer/input@5.1.9...@inquirer/input@5.2.0)

Updates `@inquirer/search` from 4.3.6 to 4.4.0
- [Release notes](https://github.com/SBoudrias/Inquirer.js/releases)
- [Commits](https://github.com/SBoudrias/Inquirer.js/compare/@inquirer/search@4.3.6...@inquirer/search@4.4.0)

Updates `papaparse` from 5.7.0 to 5.7.1
- [Release notes](https://github.com/mholt/PapaParse/releases)
- [Changelog](https://github.com/mholt/PapaParse/blob/master/CHANGELOG.md)
- [Commits](https://github.com/mholt/PapaParse/compare/5.7.0...5.7.1)

---
updated-dependencies:
- dependency-name: prettier
  dependency-version: 3.10.1
  dependency-type: direct:development
  update-type: version-update:semver-minor
  dependency-group: third-party
- dependency-name: prettier
  dependency-version: 3.10.1
  dependency-type: direct:development
  update-type: version-update:semver-minor
  dependency-group: third-party
- dependency-name: "@inquirer/checkbox"
  dependency-version: 5.3.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
  dependency-group: third-party
- dependency-name: "@inquirer/confirm"
  dependency-version: 6.4.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
  dependency-group: third-party
- dependency-name: "@inquirer/input"
  dependency-version: 5.2.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
  dependency-group: third-party
- dependency-name: "@inquirer/search"
  dependency-version: 4.4.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
  dependency-group: third-party
- dependency-name: papaparse
  dependency-version: 5.7.1
  dependency-type: direct:production
  update-type: version-update:semver-patch
  dependency-group: third-party
...

Signed-off-by: dependabot[bot] <support@github.com>

* chore(deps): preserve Node compatibility and simplify test output

Keep the existing Inquirer prompts and dependency trees because the new releases require Node 22.22.2 or 24.15.0 on those release lines. Enforce the advertised Node floors during installed-package smoke tests.

Share the JSON line generator in the dependency evidence test to keep the streamed events readable with Prettier 3.10.1.

---------

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Michael D'Angelo <mdangelo@openai.com>
2026-11-19 14:38:48 -08:00
utsav-oaiandutsav-oai e224a3314b fix(cli): hillclimb dependency investigation eval & speed (#2737)
* feat(cli): add independently reviewed dependency checks

* Compare frozen Python advisory versions as releases

* fix(cli): bind unnamed root manifests to observed services

* test(cli): evaluate dependency investigations across public fixtures

* test(sdk): bound the budget wait after the mock scan starts

* test(cli): exercise MCP evidence and missing caller access

* Bind graded citations to the approved use evidence

* test: restore dependency investigation quality controls

* fix(cli): restore dependency investigation and report compatibility

* test: grade malformed dependency baseline answers

* fix(cli): scope headless tool access to the investigator

* test(cli): lock down reviewer and external evidence tool access

* fix(cli): keep linked evidence tied to its originating call

* Reconcile application mappings and expose original evidence hashes

* test: verify selected catalog rows against original replies

* fix(cli): recover incomplete reviews and preserve selected evidence rows

* fix(cli): preserve wire evidence and verify finding citations

* Finalize from approved submissions and explain remaining uses

* fix(cli): restore coordinator recovery, independent review and report contracts

* Review withdrawals under the original dependency assessment ID

* test(cli): compare trace progress with completed search scopes

* Recheck dependency coverage after concurrent investigation changes

* Preserve exploratory scope explanations during dependency review

* feat(cli): track shared source work in dependency investigations

* Include shared source work in final coverage checks

* Require deciding evidence on the caller it establishes

* Preserve generic MCP search outcomes and scopes in dependency reports

* Share pinned source notes between dependency investigators

* Ship the source worklist and search audit modules

* refactor(cli): name the dependency coordinator for its ability to resume

* refactor(cli): simplify the tools for dependency investigations

* fix(cli): keep investigation tool arguments visible to Codex and reuse the evidence store

* fix(cli): send missing verdict citations back to the reviewer

* fix(cli): verify caller bindings and input flow across investigation steps

* fix(cli): require direct citations for asserted runtime facts on every use

* test(cli): verify failed catalog lookups without a timed source scan

---------

Co-authored-by: utsav-oai <348536218+utsav-oai@users.noreply.github.com>
2026-11-19 11:10:07 -08:00
utsav-oaiandutsav-oai 5313e2e857 feat(cli): add independently reviewed dependency checks (#2715)
* feat(cli): add independently reviewed dependency checks

* Compare frozen Python advisory versions as releases

* fix(cli): bind unnamed root manifests to observed services

---------

Co-authored-by: utsav-oai <348536218+utsav-oai@users.noreply.github.com>
2026-11-18 14:12:10 -08:00
utsav-oaiandutsav-oai 0d6b6b8626 feat(cli): collect repository evidence for dependency investigations (#2714)
Co-authored-by: utsav-oai <348536218+utsav-oai@users.noreply.github.com>
2026-11-18 10:38:46 -05:00
utsav-oaiandutsav-oai c1482dc808 feat(cli): investigate supplied dependency stacks (#2713)
* feat(cli): investigate supplied dependency stacks

* fix(cli): verify package versions for sampled stack verdicts

---------

Co-authored-by: utsav-oai <348536218+utsav-oai@users.noreply.github.com>
2026-11-17 17:14:31 -08:00
utsav-oaiandutsav-oai 78418ad7c9 feat(cli): add an opt-in foundation for dependency investigations (#2712)
Co-authored-by: utsav-oai <348536218+utsav-oai@users.noreply.github.com>
2026-11-17 15:25:25 -08:00
Kyle Brown acced5b84c feat(classify-severity): emit structured diagnostics on hosted records (#2175)
Allow hosts to opt in to classifier diagnostics on existing review and terminal frames. Preserve the default protocol for hosts that omit the diagnostics condition.
2026-11-17 14:26:56 -08:00
Michael D'Angelo b517c49ec4 docs: add a software citation for Codex Security (#2639)
Publish citation metadata and keep its version in sync with the rolling release proposal.
2026-11-17 00:00:58 -08:00
Carlos Matos 4822a9cbc4 feat(imports): import CrowdStrike FCS IaC findings locally (#2589)
* feat(imports): import CrowdStrike FCS IaC findings locally

Add a crowdstrike file reader for saved FCS CLI 4.x IaC JSON reports
(rule_detections). Each detection becomes one iac or secret finding with
typed rule, configuration and remediation details. Like Tenable, the
reader is local only and binds findings to --source-repository-url,
which must match a fetch remote of the selected checkout.

FCS writes detection paths relative to its working directory, so the
guide asks users to scan from the checkout root. Paths that are absolute
or leave the checkout are excluded. The source finding ID includes the
line, because detections of one rule on one resource can differ only by
line.

* fix(imports): import CrowdStrike reports without a Git checkout

An FCS IaC report describes files relative to the directory where the CLI
ran and names no repository, so requiring a matching Git remote did not
bind the report to anything. Import into any directory instead, and check
that every detection's file exists there. A supplied
--source-repository-url still records the repository, adds typed details
and must match a fetch remote when the directory is a Git checkout.

Detections in registry or Git modules name the calling file with a line
from inside the module. Use the call line that FCS states in
module_remediation, or no line, and include the module path in the
finding ID.

* docs(plugin): describe CrowdStrike FCS imports in the publish-findings skill

Tell Codex when to use --provider crowdstrike, how to run an installed FCS CLI from the import directory, which report to import after a nonzero exit, and to ask for a report instead of installing FCS or passing Falcon credentials.

* fix(imports): read the CrowdStrike module call line in linear time

Search module_remediation for the detection file's exact call phrase instead of a lazy regex. A crafted report could make the regex quadratic: a 2 MB module_remediation took about two minutes to read.

* fix(imports): stop CrowdStrike imports when a scanned file changed

FCS records each scanned file's SHA-256 as file_sha256. Compare it with
the file in the selected directory and stop, naming the file, when they
differ. Before, a report of another commit imported without an error and
could store lines that no longer point at the detected configuration.
The check follows file symlinks, as FCS does, and reads only regular
files.

* docs(imports): add CrowdStrike scan examples and Codex validation steps

Fix the FCS scan command: FCS needs an existing output directory whose
name has no dot. Show a remote git:: scan pinned with ?ref= and a
checkout of the same commit, and note that a //subdir URL gives paths
from that subdirectory. Add a shortened sample report, and show how to
pass imported findings to codex-security validate, for all findings or
for one scan.

* docs(imports): link CrowdStrike imports to the shared imported finding workflow

Other provider pages do not describe validation or patching. Remove the
CrowdStrike validation section and link the readback step to the shared
section on working with imported findings.

* fix(imports): keep CrowdStrike module lines in git:: reports

In a git:: report, FCS writes the detection file, its module_source and
the calling file in module_remediation with a leading slash. The reader
removed the slash from the file before it compared the three, so every
module detection lost its line. Compare them as FCS wrote them. In a
git:: scan of a public Terraform repository, 102 of 396 findings had no
line before this change and none after.

* fix(imports): record and check the repository of CrowdStrike git:: reports

A git:: report names the repository that FCS cloned, but the reader
ignored it. A report of one repository could be saved under another one
whose files were identical. For a git::https:// scan, record that
repository without its //subdirectory and ?ref=, and keep the scan path
in the report and in each finding's source data. When
--source-repository-url is given, it must name the same repository, on
import and on replay. Without it, the scanned repository gives typed
details but does not bind the checkout to a fetch remote, so clones from
self-hosted servers over SSH keep importing. Other git:: URL forms
import as before.

* fix(imports): import CrowdStrike git:: reports of URLs with a user name

When the scanned URL has a user name, FCS writes each detection file,
module source and calling file after the scanned URL instead of after a
slash, and the reader excluded every detection. Remove that prefix. FCS
removes the user name and token from the report's path and keeps them
only in flags.path, which the reader does not read.

* fix(imports): check replayed CrowdStrike locations against the native detection

A replayed preview with cleared or edited locations skipped the file
check, and saved findings without the scanned location. Rebuild the
location from the retained FCS detection, require each finding to show
exactly that location, and check that file and its hash.
2026-11-12 13:09:36 -08:00
jameshiester-oaiandMichael D'Angelo 2911786f24 feat(scanner-imports): import AWS Continuum code issues (#2619)
* feat(scanner-imports): collect AWS Continuum code issues

---------

Co-authored-by: Michael D'Angelo <mdangelo@openai.com>
2026-11-11 17:43:55 -08:00
Michael D'AngeloandCodex 92585995e1 fix(scans): resolve inventory and bulk tools outside selected sources (#2431)
* fix(scans): resolve inventory and bulk tools outside selected sources

* fix(runtime): preserve ordinary command startup timing

* fix(inventory): continue past sandbox-inaccessible search entries

* fix(runtime): match child process tool lookup in campaigns

* fix(inventory): preserve trusted tools for POSIX wrappers

* fix(scans): preserve campaign tool paths across child processes

* fix(runtime): preserve external plugin environment declarations

* fix(scans): follow intermediate Windows tool links

* fix(inventory): verify POSIX script interpreter paths

* fix(scans): bind campaign commands at subprocess launch

* fix(sdk): isolate bundled plugin discovery

* fix(scans): preserve installed shell launcher paths

* fix(sdk): package the campaign plugin locator

* fix(inventory): follow POSIX directory links before parent traversal

* fix(scans): preserve trusted Node lookup across scan launches

* fix(runtime): time the sandbox probe after tool selection

* fix(runtime): honor cancellation before launching the sandbox probe

* refactor(scans): scope source lookup to campaign executables

* Preserve Unicode reconnect notifications after integration

* fix: preserve physical Git protection roots and portable dispatcher bytes

* Retain standalone Node fallback and respect inspected SDK paths

* Exercise linked prompt and trusted Git selection across platforms

---------

Co-authored-by: Codex <codex@users.noreply.github.com>
2026-11-11 15:03:21 -08:00
a675744da4 fix(diagnostics): distinguish configuration errors and policy refusals (#1291)
* fix(cli): preserve diagnostic context and saved-scan contracts

* fix(cli): escape publication diagnostic terminal controls

* test(cli): return canonical scan IDs from recipe fixtures

* fix(cli): preserve local failure classification through causes

* test: expect requested finding page sizes

* fix(cli): keep TOML errors out of authentication advice

* fix(diagnostics): preserve typed failures and report lookup errors once

* Preserve explicit cybersecurity policy refusal handling

* fix(cli): escape C1 controls in launcher failures

* fix: escape C1 controls in CLI failure diagnostics

* fix: recognize explicit nominal policy refusals

* fix(cli): retain quota advice and escape configuration diagnostics

* fix(cli): retain quota advice and format config failures

* fix(diagnostics): escape config errors and return accepted finding pages

* fix(diagnostics): escape component planning config errors

* fix(cli): format export failure diagnostics

* test(plugin): scope safe-path helper checks to Python 3.11+

* test(plugin): exercise isolated helper imports on Python 3.10

* fix(cli): retain explicit model permission advice

* test(cli): cover public export diagnostic controls

* fix(cli): retain provider error aliases and explicit refusals

* fix(cli): retain recovery advice and Git failure status

* Recognize explicit policy refusals across response wording

* fix: retain skill error aliases and explicit policy refusals

* Keep source-access errors distinct from explicit policy refusals

* fix(cli): limit TOML exclusion to parser locations

* fix: retain explicit model and policy failure advice

* Recognize explicit policy-first review refusals

* fix(review): distinguish service errors from policy refusals

* Recognize explicit causal policy refusal responses

* fix(cli): preserve saved scan output modes

* fix(cli): retain expired token recovery advice

* Reuse executable fixture cleanup after Node launches

* fix: distinguish blocked artifact uploads from review refusals

* fix: preserve Git errors and distinguish operational refusals

* fix(review): distinguish publication failures from policy refusals

* test: use shared RPC request helper for diagnostic regression

* test: adapt ranking diagnostics to shared repository CLI fixture

* fix(cli): retain advice for token expiry code variants

* fix(review): recognize explicit policy accordance refusals

* fix(cli): preserve advice for canonical diagnostic codes

* fix: preserve operational failures and escape terminal controls

* fix(cli): require status context for numeric failure advice

* fix(cli): retain quoted status and terse refusal handling

* fix(cli): retain explicit HTTP code diagnostic advice

* Fix Windows direct Deep package fixture spawning

* Keep technical request failures retryable

* Retain parser context and anchor policy-prefix refusal messages

* Preserve explicit refusal sentences with explanatory text

* Preserve same-sentence policy refusal explanations

* fix(diagnostics): preserve refusal sentence boundaries

* fix(diagnostics): retain complete policy refusal clauses

* fix(diagnostics): preserve refusal sentences and transient retries

* test: await tar extraction in package fixtures

* fix: recognize unsafe-content review refusals

* test: preserve provider diagnostics in CLI assertions

* fix: preserve explicit colon-form review refusals

* fix: exclude URL context from diagnostic classification

* fix(contract): preserve exact checked file identities

* fix: recognize descriptive content refusal reasons

* fix(cli): exclude endpoint URLs before TOML diagnostics

* fix: classify explicit policy refusals independently of reasons

* fix(cli): recognize separated response-code statuses

* fix: preserve structured review transport failures

* fix(diagnostics): keep operational policy failures distinct

* fix(cli): retain advice for named model failures

* fix(cli): keep local timeout diagnostics distinct

* fix: mark canceled patch review output as unsuccessful

* fix(cli): retain advice for response status fields

* fix(cli): keep local model errors free of credential advice

* fix(cli): retain saved results in full YAML and TOON failures

* fix(diagnostics): preserve advice for quoted provider errors

* test: forward directory probe arguments

* test(cli): use the existing artifact export import

* test(sdk): assert native scan artifact diagnostics

* ci: retry managed Code Quality after GitHub installation rate limit

* Keep local diagnostics distinct and surface inaccessible issue history

* Remove duplicate coordinator test import inherited from main

* Retain the SQLite diagnostic alongside the typed inspection exit

* Verify the printable path escaping used by export errors

* Classify long skill diagnostics without rescanning each suffix

* Preserve message structure when masking qualified model names

* Preserve backticks around provider-qualified model names

* Recognize standalone safety and content policy refusals

---------

Co-authored-by: Codex <codex@users.noreply.github.com>
Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-11 12:20:12 -08:00
Michael D'Angeloanddy-armorcode 5aa4d394ea feat(scanner-imports): add ArmorCode findings support (#2608)
Carry forward the reader contributed in #2606. Include it in the SDK package and validate repository claims when preparing an import.

Co-authored-by: dy-armorcode <109572335+dy-armorcode@users.noreply.github.com>
2026-11-11 11:17:28 -08:00
Michael D'Angelo 5811e2df3d feat(imports): publish versioned workbench findings through the shared Cloud transport (#2603)
* feat(imports): publish versioned workbench findings through the shared Cloud transport

* fix(imports): preserve acknowledged batches when Cloud rejects a later batch

* test(workbench): restore the v53 schema before replaying import migrations

* fix(imports): keep versioned SQLite selections alive through readback

* fix(workbench): retain version-scoped Cloud receipts across repository resets

* Resume Sonar Cloud retries from the stored import checkpoint

* Keep the Sonar replay assertion strictly typed
2026-11-11 10:47:31 -08:00
Michael D'Angelo c1730abab3 fix(cli): preserve JSON errors and token windows (#2593)
* fix(cli): make saved scan selection and matching explicit

* fix(cli): resolve command paths and isolate standalone actions

* fix(cli): preserve updater result and diagnostics

* fix(cli): simplify finding imports and protect export files

* fix(cli): preserve JSON output on failures and token windows

* fix(cli): identify missing global option values before routing

* fix(package): include the JSON token view in the SDK distribution

* test(plugin): initialize Windows registration fixtures explicitly

* test(cli): distinguish the legacy publish check selector in help

* test(cli): check the requested JSON format in raw CSV diagnostics

* fix(cli): retain suggestions for unknown commands

* test(cli): exercise late Markdown format selection with valid global options

* style(cli): format argument routing regressions

* style(cli): format structured output regressions

* test(cli): return worktree selection in authentication fixture

* fix(cli): reject missing global values before command dispatch

* fix(cli): preserve short help before a command

* fix(cli): reject switches as global option values

* fix(cli): identify package updates in generated agent guidance

* fix(cli): serialize implicit help and completion scripts for JSON output
2026-11-11 18:38:13 +00:00
Michael D'Angelo 1e3c82abe5 refactor(imports)!: retire historical reader formats and Cloud checkpoints (#2602)
* refactor(imports)!: retire historical reader formats and Cloud checkpoints

* test(imports): remove an empty reader import

* refactor(imports): project saved checkpoint items once
2026-11-11 10:00:11 -08:00
Michael D'Angelo ae291b2af3 fix(cli): validate inputs and make plans and hooks reusable (#2594)
* fix(cli): validate local inputs and make saved plans and hooks reusable

* fix(cli): bound batch scan spending and follow-up patches

* fix(cli): label total cost arguments in USD

* fix(cli): carry scan overrides into component plan commands

* fix(package): include the batch-cost module in the published SDK

* fix(cli): preserve local path and budget settings during preflight

* docs(cli): format policy and time-limit command references

* test(cli): pass Windows drive path explicitly as literal text

* fix(cli): close hook handles before replacing files on Windows
2026-11-11 17:14:58 +00:00
Michael D'Angelo 6b0d0a7ce1 fix(cli): honor scan budgets across batch attempts and follow-ups (#2592)
* fix(cli): bound batch scan spending and follow-up patches

* fix(cli): label total cost arguments in USD

* fix(package): include the batch-cost module in the published SDK

* fix(cli): skip unmetered owner suggestions under a total budget
2026-11-11 05:56:29 -08:00
Michael D'Angelo 721cfce630 fix(cli)!: make MCP targets and launchers explicit (#2591)
* fix(cli)!: make agent installation targets and launchers explicit

* fix(cli): parse positional arguments after MCP launcher paths

* fix(cli): detect the configured Codex profile for MCP installation

* fix(package): include the command-words module in the published SDK

* chore(package): group the command parser with cost configuration

* fix(cli): preserve Windows paths in quoted MCP launchers
2026-11-11 05:20:30 -08:00
ea617a4c09 fix(workbench): retain shared history across worktree changes (#456)
* fix(workbench): share scan history across linked worktrees

* fix(workbench): verify repository generation across platforms

* fix(workbench): harden and simplify worktree history

* fix(workbench): preserve recreated repository subdirectories

* fix(workbench): refresh the bundled history plugin

Advance the paired standalone plugin version so existing staged marketplaces load the linked-worktree history helpers. Leave the npm package release version unchanged.

* fix(workbench): verify saved history candidates

Use recorded checkout ownership before falling back to live repository identities. Preserve matching persisted identities for removed worktrees and retain verified same-origin comparisons.

* fix(workbench): unify verified repository history

* fix(workbench): preserve scan history ownership and replay

* fix(workbench): quarantine unproved prerelease identities

* fix(workbench): order automatic history comparisons

* fix(workbench): persist completion visibility order

* fix(workbench): preserve findings and Linux identities

* fix(workbench): preserve unbound legacy history

* Bind repository history to each scan generation

* Preserve archived scans across registration failures

* Bound repository identity maintenance to selected targets

* Repair legacy identity upgrades and precompute repository counts

* test(workbench): isolate directory metadata on Python 3.14

* fix(sdk): exclude internal helpers from public declarations

* test(workbench): send identity fixture through stdin

* fix(workbench): verify ownership before rejoining scans

* test(workbench): retain recreated fixture inodes

* test: align shared history fixtures with current schema

* test: include current migrations in repository identity upgrade checks

* Preserve owned legacy history and current rerun prompts

* Preserve repository history through validated saved matches

* Preserve timestamp and custom plugin history compatibility

* Preserve grouped history matching and mock output recovery

* Preserve grouped history matching and mock output recovery

* Preserve portable repository ownership and indexed history lookups

* Preserve legacy Git history and mixed filesystem comparisons

* Preserve legacy triage and registered linked-worktree comparison

* Keep legacy comparisons scoped and reuse repository history indexing

* Rank same-target feedback across legacy repository generations

* Keep legacy history local when matching current repository targets

* Preserve saved legacy aliases in linked comparison history

* Order combined feedback fixture imports

* Update completed findings fixture for combined migration history

* Align migration expectations and batch workflow condition fixtures

* Adapt archived-output callbacks to current observer helper

* Format archived-output observer callback

* Use the current historical migration fixture helper

* Preserve requested repository aliases through findings lookup

* Preserve local legacy decisions and requested scan aliases

* Support absolute paths in the Windows history fixture

* test: preserve Windows fixture identity and shell behavior

* fix(history): preserve accepted completion timestamp precision

* fix(workbench): replace reversed scan comparisons

* docs: make finding detail example explicitly synthetic

* test(workbench): follow shared scan allocation in identity fixture

* fix(workbench): preserve history across filesystem remounts

* test(workbench): align identity fixtures with migration semantics

* fix(workbench): preserve history across supported Python runtimes

* perf(workbench): paginate repository projections

* test(workbench): type repository page expectations

* fix(workbench): retain verified history after device changes

* test(workbench): supply archive defaults in lineage fixture

* fix(workbench): preserve composed remediation and writer fixtures

* Preserve caller scope and valid history during saved scan bootstrap

* Retain bound history after worktree paths are reused

* Use scan generations for history and live Git roots for interpreter protection

* Align schema and archive rollback fixtures with registration ordering

* Exclude saved checkouts from ambient Python discovery

* Isolate trusted Python from Git in history fixture

* Preserve base Python and packages in isolated history fixture

* test(findings): expect the requested repository alias

* fix(identity): avoid external stat and synchronous fixture timeouts

* fix(workbench): preserve historical target registration and migration

* fix(history): compare saved scan timestamps chronologically

* fix(sdk): trim saved scan timestamp fractions linearly

* Fix platform coverage in history migration fixtures

* fix(workbench): retain lexical roots during migration discovery

* test: preserve shared history across receipt migration

* test: align repository generation fixtures with merged migrations

* test(workbench): align identity migrations with issue receipts

* test(workbench): align history fixtures with current contracts

* test: run migration probes through Python runtime helpers

* Update generation migration fixtures for current schema

* test(workbench): align coordinator fixture with generation migration

* fix(workbench): keep Git discovery scoped across saved history

* fix(workbench): cover new and unavailable Git targets

* fix(workbench): expand incoming Git target paths

* docs: format shared scan history guidance

* fix(workbench): retain evidence of vanished Git targets

* test: exercise current repository generation migration

* fix: preserve Git discovery after incomplete workspace setup

* test: bind host Git in direct saved-history helpers

* fix(workbench): retain all scan feedback and register fixture targets

* fix(workbench): preserve saved checkout boundaries during tool discovery

* test(plugin): restore checkout root column in legacy migration fixture

* fix(plugin): require recorded checkout roots for legacy Git discovery

* fix(sdk): require saved checkout roots before tool discovery

* test(workbench): install the lifecycle view in the repository alias fixture

* fix(sdk): exclude rejected workspace paths from executable discovery

* fix(workbench): retain trusted Git for deleted bounded targets

* fix(workbench): disable ambient lookup for unanchored workspace paths

* fix(workbench): retain checkout boundaries for rejected workspaces

* test(workbench): use the complete schema in Python migration fixtures

* Retain rejected checkout exclusions after workspace correction

* Classify the quota error in the comparison stop test

* Update workbench identity fixtures with the object-store birth time

* Exercise repository generation upgrades and Windows owner rejection after renumbering

* Drop the stale retry assignment after archive acknowledgement

* Start the Windows archive attempt when the real patch write opens

* Exercise repository history with the current registration and schema

* Keep workbench identity test outputs in private temporary directories

* test: verify current and stale checkout ownership independently

* fix(history): validate native parent traversal before lookup

* Cover partially sequenced completion history on migration replay

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
Co-authored-by: Codex <codex@users.noreply.github.com>
2026-11-11 04:46:42 -08:00
Michael D'AngeloandCodex 83c9c9196b fix(cli): inspect integration commands before installation (#2582)
* fix(cli): inspect integration commands before installation

* fix(cli): include the integration dispatcher in package validation

* fix(cli): list generated skills at the requested grouping depth

* docs(cli): clarify which installation skills list checks

* refactor(cli): share integration commands across dispatch and discovery

* fix(cli): complete integration flags without executing their commands

* fix(cli): register MCP clients through the installed SDK dependency

* fix(cli): preserve shell completion words and cursor positions

* fix(cli): serve all shell completions before executable dispatch

---------

Co-authored-by: Codex <noreply@openai.com>
2026-11-11 01:41:00 +00:00
Michael D'Angelo facaa9e07d feat(cli): expose identifiers for saved scan follow-up (#2577)
* feat(cli): expose identifiers for saved scan follow-up

* test(sdk): exercise dedupe representatives in publication

* feat(cli): link interrupted scans and imported evidence to their next commands

* build(sdk): include the command argument helper in package checks

* fix(cli): scope recovery hints to supported sessions and shells

* refactor(cli): simplify the result handoffs and verify PowerShell recovery
2026-11-11 01:40:40 +00:00
Michael D'AngeloandMichael D'Angelo 9534e7fd6e docs: shorten the npm package README (#2542)
* docs: shorten the npm package README

* docs: restore scan scope, CI behavior, and results guidance

---------

Co-authored-by: Michael D'Angelo <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 21:40:18 -08:00
Michael D'Angeloandmldangelo-oai 3254a21592 fix(scans): archive long POSIX output directory names (#2529)
* fix(scans): archive long POSIX output directory names

* fix(logs): recover bound sessions from short scan archives

* fix(package): include the scan archive path helper in the release contract

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 20:21:32 -08:00
Michael D'Angeloandmldangelo-oai 0e606bf1bc fix(package): explain the Alpine tar prerequisite in release checks (#2536)
Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 20:19:50 -08:00
Michael D'Angeloandmldangelo-oai f3efffee06 fix(package): build the bundled plugin in long Windows checkouts (#2508)
* fix(sdk): build bundled plugins from long Windows paths

Use the extended Windows path only when creating the temporary build directory. Keep the selected parent spelling for the compiler, publication, and cleanup, and exercise the package build from a long source snapshot.

* fix(plugin): build the selected MCP app on long Windows paths

Run the declared app build with its local dependency binaries when the Windows Node task runner would skip a long package or tool path. Keep the native runner on shorter paths, other platforms, and Bun.

* fix(plugin): run the selected Bun build script in deep Windows checkouts

Use Bun run build for long Windows MCP app paths so the reserved bundler command cannot shadow the package script. Exercise the real Bun process and its selected package lifecycle in the long source-build regression.

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 18:07:06 -08:00
Michael D'Angeloandmldangelo-oai 025fd5be3e fix(package): prevent stale SDK archives on long Windows paths (#2511)
* fix(pack): stop before Windows runs an ancestor SDK build

* Keep prepack boundary tests within the available test directory

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 16:51:06 -08:00
Michael D'Angeloandmldangelo-oai 8716bbb373 fix(sdk): finish plugin builds while old Windows addons are loaded (#2502)
A mapped native addon can survive both directory renames and block deletion of the retired bundle. Retry its removal within the existing Windows file-lock budget. If readers still hold it, return the validated replacement and report the invocation-specific staging directory for cleanup.

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 14:06:09 -08:00
Michael D'Angeloandmldangelo-oai 88d6b757df fix(windows): retry publishing validated plugin ZIPs (#2498)
* Retry custom plugin ZIP publication on Windows

* fix(sdk): declare the Windows rename helper for source consumers

Strict eval builds type-check the SDK source without JavaScript inference. Pair the shared helper with its source declaration and include its JavaScript explicitly so clean SDK builds still emit the implementation, declarations, and maps.

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 13:57:36 -08:00
Michael D'Angeloandmldangelo-oai 9ee5983b30 test(sdk): resolve a usable Python interpreter for Windows workbench tests (#2493)
* test(sdk): resolve a usable Python interpreter for Windows workbench tests

* fix(package): include the plugin Python probe in the dist inventory

* test(sdk): prepare Windows Python outside individual test deadlines

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 13:49:54 -08:00
75633fb5fd fix(scans): warn when repository HEAD changes during a scan (#360)
* fix(sdk): warn when repository HEAD drifts

Signed-off-by: Mr-Neutr0n <harikp2002@gmail.com>

* fix(targets): stop HEAD monitoring at scan completion

* fix(scans): stop HEAD monitoring before failure follow-ups

* fix(scans): keep early HEAD drift warnings informational

* fix(packaging): include target drift module in package checks

* refactor(scans): schedule HEAD checks after the previous read settles

---------

Signed-off-by: Mr-Neutr0n <harikp2002@gmail.com>
Signed-off-by: Harikrishna KP <harikp2002@gmail.com>
Signed-off-by: Mr-Neutr0n <64578610+Mr-Neutr0n@users.noreply.github.com>
Co-authored-by: Cursor Agent <cursoragent@cursor.com>
Co-authored-by: Michael D'Angelo <mdangelo@openai.com>
Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 13:45:19 -08:00
Michael D'Angeloandmldangelo-oai ea8c16644f fix(package): preserve LF in the POSIX MCP launcher on Windows (#2500)
Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 13:25:52 -08:00
Michael D'Angelo 61fc3a061f fix(release): sign macOS addons for packaged review items (#2331)
* fix(release): sign macOS addons for packaged review items

* fix(release): bind signing to the verified release artifacts

* chore(ci): retry GitHub Code Quality after upload rate limit

* fix(release): sign Darwin addons through Signer Service
2026-11-09 08:37:24 -08:00
Michael D'Angeloandmldangelo-oai 884b7b657a refactor!: consolidate scan, workbench and release rules (#2452)
* refactor(workbench): share schema repair and saved finding projections

* refactor(cli): resolve the command route once

* refactor(settings): share Deep Scan configuration and restore saved inputs

* refactor(imports): reuse scanner projections at the file boundary

* refactor(publish): share pending request and receipt handling

* refactor(telemetry): classify publication failures in the recorder

* refactor(package): share plugin inventory and archive checks

* refactor(release): reuse verified provenance and release decisions

* fix(ci): reconcile release labels with an empty array on Bash 3.2

* ci: report failed source checks after cancelled PR edits

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-09 05:02:21 -08:00
Michael D'Angelo 7607a4d504 refactor(cli): share terminal error listener cleanup (#2481)
Progress, the dashboard and command scopes each installed an optional
stderr error listener and deferred its removal through the final write.
The dashboard checked only whether its timer was stopped. A previous
flush could remove the listener after a restart while later output was
still pending. Reuse Progress's generation check across all three and
include the new module in the package inventory.

A delayed error after two dashboard starts and stops exited 1 on
6e677445 and the published 0.8.3 CLI, and 0 with this change. Fifty
blocked progress cycles kept identical terminal and piped output across
Node 22, 24 and 26. The focused regression, 574 CLI tests, SDK types,
formatting, and the 955-file package inventory passed. The extracted
package also passed launcher and renderer checks.

The full installed SDK smoke and 19-scenario PTY suite passed on the
preceding build with identical dashboard and listener JavaScript. Runs
on the final tree reached subprocess deadlines under concurrent load,
including the installed plugin compatibility fixture. The final tree's
CLI JavaScript differs from that build only in import ordering.
2026-11-09 04:08:17 -08:00
Michael D'Angelo 5b1e8a53de refactor(release): resolve history in the GitHub publisher (#2475)
Read published versions, release tags, and reachable tags in the publisher before generating notes. Keep the previous tag and latest decision in their native types and remove the intermediate workflow outputs. The internal release-history command had one workflow caller. No public API or client migration is required.

Verified with 318 release automation and release progress workflow tests. Thirteen synthetic comparisons of the previous and revised pipelines matched command arguments, generated notes, and exit codes for initial releases, backfills, partial histories, and failures.
2026-11-09 04:07:32 -08:00
Michael D'Angelo 03915c1fd7 fix(sdk): generate project schemas before bundling MCP (#2473)
* fix(sdk): generate project schemas before bundling MCP

Generate plugin-derived SDK defaults and their dependent JSON schemas in a single ordered build step. Full package builds run this step before the plugin snapshots the schema. The projection-only build:plugin command continues to work from a sparse source snapshot. Published CLI, SDK, and schema paths need no migration.

This removes 11 maintained build lines and 8 lines overall. An offline pack with a temporary worker cap previously let the embedded MCP accept nine workers although the SDK rejected it. Both now reject nine and accept eight in the actual packed artifact. On current inputs all 951 package entries and the schema/default artifacts matched the baseline. The targeted suites passed 90 tests with three Windows-specific skips on Linux.

* docs(sdk): describe generation beside the pack command

Keep the pack sequence with the command that runs it. The standalone plugin section can describe its separate projection and payload layout.
2026-11-09 04:07:19 -08:00
Michael D'Angelo f956b5792f refactor(release): retire first-publish npm bootstrap (#2472)
The public package has already shipped 0.1.0 and npm permanently reserves
that name and version. Remove the first-publish token step and the local
adapter that treated an E404 as an empty release history. Both workflows
stop on lookup failure. Normal publication still uses trusted publishing.
Historical GitHub release verification keeps its original recovery path.

Across 22 synthetic release-cut and publisher comparisons, only the two
initial-404 cases changed. The candidate passed 407 release, Action, and
container tests. The historical 0.1.0 npm and GitHub archives matched
byte for byte. The pinned npm signature audit and both revisions'
provenance and publication checks accepted the same archive and run.
No package user or downstream release command needs to change.
2026-11-09 04:07:11 -08:00
Michael D'Angelo ce25486e8d refactor(package)!: reject sparse entries in release tarballs (#2456)
Remove sparse reconstruction from the checkout's package checker. Refuse
sparse metadata before native extraction. Check payload sizes and bytes
against the native tar view and keep the independent npm content checks.

Maintainers with an unsigned candidate produced by a sparse-aware tar
tool can rebuild it from the same source with pnpm pack. Keep exact
published or signed archive bytes for release recovery. Installed npm
CLI and SDK consumers require no migration.

All 42 historical npm releases (16,383 entries) matched the prior reader
for paths, sizes, data, metadata, npm interpretation and disclosure
checks. Both npm and pnpm packed a file with holes as an ordinary entry.
Four GNU sparse encodings now reject that input explicitly.

Passed 82 package tests, SDK types and formatting, nine plugin source
compatibility tests, and the full installed-package smoke for the normal
951-entry tarball on Linux. Local checks used GNU tar and the libarchive
reader. Native Windows and macOS execution remains for CI.
2026-11-09 04:04:18 -08:00
khai-oaiandMichael D'Angelo 794edbe957 chore(deps): update the Bun toolchain to 1.4.4 (#2417)
* ci(deps): update Bun to 1.4.4

* test(ci): align the Bun workflow version assertion

* chore(deps): align remaining Bun toolchain pins

* fix(ci): preserve test timeouts with Bun 1.4 options

---------

Co-authored-by: Michael D'Angelo <mdangelo@openai.com>
2026-11-06 18:38:02 -08:00
Michael D'Angeloandmldangelo-oai dc80da644e fix(windows): open long state paths in JavaScript SQLite connections (#2380)
Choose a Windows filename at the five file-backed Node and Bun SQLite opens. Cover credential locking, workbench initialization, and saved-history reads while preserving the native byte limit and public path spelling.

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-06 07:28:25 -08:00
Michael D'Angeloandmldangelo-oai 206a9c8106 fix(build): retry Windows plugin bundle renames (#2370)
Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-06 06:44:31 -08:00
Michael D'Angelo 37c1a5fd34 fix(sdk): use published Vertex AI prices for managed OpenAI models (#2363)
* fix(sdk): use published Vertex AI prices for managed OpenAI models

* fix(package): include Vertex pricing in the SDK manifest
2026-11-05 16:39:09 -08:00
Michael D'Angelo 31a42e9ba7 fix(sdk): price OpenAI models by Bedrock region and inference route (#2362)
* fix(sdk): price OpenAI models by Bedrock region and inference route

* fix(package): include Bedrock pricing modules in the SDK manifest
2026-11-05 16:09:45 -08:00
Michael D'AngeloandMichael D'Angelo 0ed54be43a feat(imports): enable external scanner imports by default (#2360)
* feat(imports): enable external scanner imports by default

Register collection, publication, and readback commands directly. Keep provider credential checks, repository access checks, and upload confirmation in their existing paths. Clients can remove the retired environment variable.

* docs(imports): make scanner provider guides evergreen

Index the scanner report readers and supported destinations. Use installed CLI examples, distinguish repository helpers from packaged tools, and describe Cloud verification and retry steps from the preview and receipt.

* docs(integrations): clarify connector setup and publication paths

* docs(integrations): simplify publication history and storage guidance

* docs(triage): specify how to exhaust Jira search results

---------

Co-authored-by: Michael D'Angelo <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-05 15:30:02 -08:00
Michael D'Angeloandmldangelo-oai 6565fd2b1e feat(imports): import Tenable One code repository findings locally (#2343)
* feat(imports): import Tenable One code repository findings locally

* refactor(tenable): make download retry failure explicit

* refactor(imports): simplify Tenable export and local import

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-05 10:17:10 -08:00
Michael D'Angelo 4d83ce9195 refactor: share workbench arguments and simplify SDK and eval workflows (#2325)
* Simplify workbench dispatch, SDK helpers, and their test fixtures

* test(sdk): keep distinct progress matrix helpers

* test(sdk): expect canceled lifecycle for aborted component scans
2026-11-05 01:28:44 -08:00