31 Commits
Author SHA1 Message Date
Michael D'Angelo ed649f6337 fix(action): preserve literal scan paths (#2617)
* Preserve literal Linux paths in the GitHub Action

* Honor completed-scan path rules in the Action
2026-11-11 15:04:22 -08:00
Michael D'AngeloandCodex 6b17ee2d0f fix(cli): update the active installation and migrate legacy options (#2579)
* fix(cli): update the active installation and migrate legacy options

* fix(cli): recognize local projects named lib during updates

* fix(cli): use supported pnpm workspace and Bun optional flags

* fix(cli): isolate self-updates from project configuration

* fix(cli): recognize global npm under nvm checkouts

* refactor(cli): simplify package manager update resolution

* refactor(cli): skip updater registration during inspection

* refactor(cli): share metadata flags with update notices

---------

Co-authored-by: Codex <noreply@openai.com>
2026-11-11 01:40:50 +00:00
Michael D'Angelo 0497da5987 fix(action): show scan scope limitations in the job summary (#2435) 2026-11-09 04:17:16 -08:00
Michael D'Angelo f9fb15c393 test(action): check declared inputs against parser reads (#2457)
Remove the separate nineteen-name registry. Compare action.yml with the keys requested by the parser for default, external-provider, and opt-in Deep Scan inputs. Use the declared names to check README examples and isolate test environments.

Verified with npm run validate (223 tests, unchanged distribution), six synthetic mutations including unconditional and conditional parser typos, and six keyless CLI preflights against 0.6.0 and 0.8.3. Action input syntax and defaults are unchanged; no migration is required.
2026-11-09 04:05:08 -08:00
Michael D'Angelo 2d50d7aff8 chore(deps): reduce the dependency tree for evals, SDK and tooling (#2409)
* chore(deps): reduce the dependency tree for evals, SDK and tooling

* docs(evals): explain how to enable optional Promptfoo peers
2026-11-06 18:08:28 -08:00
wj 07c88c1bf3 fix(action): use the singular for one additional finding in the job summary (#2252) 2026-11-03 14:13:37 -08:00
Michael D'Angeloandmldangelo-oai f601cd659b feat(action)!: prepare hosted runners and require completed PR checks (#1580)
* fix(sdk): restore GPT-6 Sol cost estimates and budgets

* feat(action)!: prepare hosted runners and require complete CI reviews

* fix(action): handle containers and retargeted pull requests

* Align the main Action quick start with sandbox and coverage behavior

* test(packaging): await tar extraction in package fixtures

* fix(action): preserve required scans across metadata edits

* fix(action): verify sandbox startup before hosted setup

* fix(action): only provision Ubuntu after sandbox startup failures

* fix(action): preserve setup diagnostics and check native configuration

* fix(action): preserve runner setup and PR check semantics

* fix(action): verify live PR targets on the triggering host

* docs(action): clarify check migration and SARIF permissions

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-11-02 18:23:46 -08:00
andy-tang-oaiandAndy Tang eeb969e87b fix(action): show highest-severity findings first (#2160)
* fix(action): show highest-severity findings first

* fix(action): preserve CLI finding order in scan results

---------

Co-authored-by: Andy Tang <332202880+andy-tang-oai@users.noreply.github.com>
2026-10-30 14:22:54 -07:00
wjandmldangelo-oai af94f47318 fix(action): do not split surrogate pairs in truncated report text (#2133)
* fix(action): do not split surrogate pairs in truncated report text

* fix(action): preserve unpaired surrogates in report text

---------

Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-10-29 15:48:29 -07:00
12 df82a6f575 fix(action): check RUNNER_TEMP before resolving it (#2125)
The action called realpath on RUNNER_TEMP before checking that it was
set, so a missing value failed with "ENOENT: no such file or directory,
realpath ''" and the "RUNNER_TEMP is required." error was never
reached. Check it first.
2026-10-28 23:31:44 -07:00
ianw-oai 621bad47e6 refactor!: simplify scan completion and recovery (#1762) 2026-10-26 17:19:54 +00:00
Michael D'Angeloandmldangelo-oai b3e9d40f5a test(tooling): reduce repeated Action and eval setup (#2040)
Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
2026-10-26 22:06:56 +05:30
Michael D'Angelo 3dfed19297 test: reuse triage replay builds and exercise native Action SARIF (#1977) 2026-10-25 19:10:49 +08:00
Michael D'Angelo 1b3eab1a0b fix(action): validate generated files before rebuilding them (#1934) 2026-10-25 09:09:28 +08:00
Michael D'Angelo e53b6c0757 ci: require checks for changed Action and eval components (#1931) 2026-10-25 08:51:13 +08:00
Michael D'Angelo 38a31ff648 feat(action): report usage and extend integration attribution (#1703) 2026-10-24 01:27:42 +08:00
Michael D'Angelo e10e28113f fix(action): isolate artifact upload failures (#1719) 2026-10-24 01:27:21 +08:00
Michael D'Angelo 859c6b2701 test(action): capture expected archive failure diagnostics (#1795)
Capture expected archive failure diagnostics in action tests.
2026-10-20 12:08:58 +08:00
Michael D'Angelo 2c6028b23b fix(telemetry): preserve controls and share request attribution (#1691) 2026-10-20 10:28:15 +08:00
Michael D'Angelo 5b4a553e24 feat(action): support GitHub Enterprise Cloud on ghe.com (#1713)
Support GitHub Enterprise Cloud checkout origins and host-aware routing, share Git remote parsing between the Action and SDK, and preserve packaged modules and GitLab IPv6 authorities.

Integrate main while retaining its Action dependency patches, publication changes, and validation fixes.
2026-10-20 06:59:58 +08:00
Michael D'Angelo c6d0bf8d78 fix(action): propagate artifact stream failures and preserve diagnostics (#1671) 2026-10-20 04:43:13 +08:00
Michael D'Angelo 7d3325abce fix(config): pass native settings through scan workers (#1717) 2026-10-20 04:39:51 +08:00
wjandMichael D'Angelo f61dfd9fe7 fix(action): describe positive numeric input bounds (#1605)
Co-authored-by: Michael D'Angelo <mdangelo@openai.com>
2026-10-19 15:15:36 +08:00
Michael D'Angelo 009228d09c feat(action): support scan CLI inference providers (#1677)
* feat(action): support native Amazon Bedrock scans

* feat(action): align provider inputs with scan CLI

* fix(cli): allow credential-free provider previews
2026-10-17 11:14:34 +09:00
Michael D'Angelo c380befb9e fix(scans)!: simplify scan outcomes (#1633)
Simplify scan outcomes across the CLI, SDK, and GitHub Action. Completed report-only scans return 0, an opt-in findings threshold returns 1, and execution or required-output failures return 2. Remove partial and unknown outcomes while retaining coverage diagnostics and saved results.

Require the current plugin for Deep Scan and canonical fields for new findings. Preserve readability of historical sealed reports. Update documentation and tests, and remove obsolete compatibility paths.
2026-10-14 11:32:39 -07:00
Michael D'Angelo 68ddbdce0c test: share completed scan fixtures (#1620)
* test: share completed scan fixtures

* test: run Action CI for shared scan fixture changes
2026-10-13 11:50:23 -07:00
mldangelo-oai eaed287d65 feat(release): align Action publication with CLI releases (#1418)
* feat(release): publish matching Action after verified CLI release

* fix(release): isolate Action verification from publication

* fix(release): verify Action runtime through Socket

* fix(ci): shard isolated Linux test runs

* fix(release): validate reused Action tags before installation

* test(release): keep Socket exit cases distinct in reports

* ci: pin Linux quality tests to the Bun subprocess fix
2026-10-09 08:17:47 -04:00
dans-oaianddans-theaicompany 8ff5b39e1e feat(action): support optional Cyber access program selection (#1516)
* Support optional Cyber access program in GitHub Action

* Document Daybreak access and refusal failures in Action

* Clarify Daybreak quickstarts and Ubuntu sandbox setup

---------

Co-authored-by: dans-theaicompany <dans@theaicompany.com>
2026-10-08 14:02:49 -03:00
mldangelo-oai f82ae5be89 fix: simplify action cleanup and stabilize provider isolation tests (#1421)
* fix: simplify action cleanup and resolve code quality warnings

* test: remove repeated native startup from provider isolation checks
2026-10-07 22:42:38 -04:00
mldangelo-oaiandCodex da446104d4 refactor(action): derive inputs and consolidate test configuration (#1348)
Co-authored-by: Codex <codex@users.noreply.github.com>
2026-10-06 23:55:13 -04:00
0cdb22a6db feat(action): add Codex Security GitHub Action (#1015)
* feat(action): integrate GitHub Action alongside the CLI

* fix(action): update pinned CLI to 0.1.29

* fix(action): omit estimated cost from scan summaries

* docs(action): simplify setup and usage guide

* docs(action): schedule weekly scans in quick start

* docs(action): minimize quick-start scan configuration

* docs(action): configure Luna PR scans and verify annotations

* Update README.md

* fix(action): update pinned CLI to 0.1.30

* fix(action): normalize paths and preserve results on summary failure

* refactor(action): remove fixed-value public inputs

* refactor(action): use job check with explicit scan outcomes

* refactor(action): use runtime manifest as single CLI version source

* fix(action): explain incomplete scan coverage in summaries

* docs: add GitHub Actions quick start to main README

* Update README.md

* fix(action): defer policy and knowledge-base handling to CLI

* refactor(action): consume validated CLI results

* fix(action): preserve scan outcomes when SARIF is unavailable

* fix(action): allow configured Dependabot scans

* test(action): add hosted dry-run smoke job

* fix(action): ignore unrelated input environment variables

* refactor(action): simplify runtime setup

* refactor(action): simplify inputs for repository and PR scans

* chore(action): update Codex Security CLI to 0.1.31

* feat(action): restore deep scan mode and time budget

* fix(action): warn on valid partial scans

* fix(action): refresh audited dependencies and preserve diagnostics

* fix(action): escape legacy runner commands in diagnostics

* fix(action): preserve result and SARIF reporting semantics

* fix(action): keep report text and selected paths literal

* fix(action): accept existing literal scope filenames

* fix(action): check large diffs without capturing path lists

* test(action): wait for child readiness before timeout

* refactor(action): delegate scanner behavior to the CLI

* fix(action): preserve runner tools and bind scan reports to the requested commit

* fix(action): preserve runner tool selection and process cleanup

* fix(action): clean up descendants before draining process output

* fix(action): preserve Python loading and inspect worktree configuration

* fix(action): preserve runner tracking through MCP launchers

* test(plugin): preserve interpreter loader settings in worker fixtures

* test(action): create the CLI integration runtime temp directory

---------

Co-authored-by: dans-oai <321942207+dans-oai@users.noreply.github.com>
Co-authored-by: mldangelo-oai <269034524+mldangelo-oai@users.noreply.github.com>
Co-authored-by: Michael D'Angelo <mdangelo@openai.com>
2026-10-05 23:50:02 -04:00