Skip to content

feat: support model-native Deep Scan workers - #270

Open
mldangelo-oai wants to merge 6 commits into
mdangelo/codex/preserve-url-scan-contextfrom
mdangelo/codex/model-native-deep-workers
Open

feat: support model-native Deep Scan workers#270
mldangelo-oai wants to merge 6 commits into
mdangelo/codex/preserve-url-scan-contextfrom
mdangelo/codex/model-native-deep-workers

Conversation

@mldangelo-oai

Copy link
Copy Markdown
Collaborator

Summary

  • Enable model-native Deep Scan workers and update the associated preflight and skill guidance.
  • Remove outdated capability-profile assumptions without changing the packaged plugin contract.

Stack

  • Intentionally stacked on mdangelo/codex/preserve-url-scan-context so this review contains only its atomic upstream change.
  • Prerequisite for the pending 0.1.6 release in release: bump Codex Security to 0.1.6 #244; package version is unchanged.

Verification

  • Parsed every bundled workbench Python module.
  • Verified all 105 declared plugin payload files exist.
  • Initialized a fresh private workbench SQLite database successfully.
  • Passed Prettier for all 2 changed formatted files.
  • Passed the focused Deep Scan ownership/continuation regression suite.
  • The complete reconstructed sync passed 775 tests (5 platform/integration skips) and the installed-package smoke test.

@mldangelo-oai

Copy link
Copy Markdown
Collaborator Author

@codex review

Please review the exact current head 205fb5d2c1a7da3a51522754356e101fec4138a9 and only this atomic PR's declared base-to-head diff.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 205fb5d2c1

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Some remediation patches have `kind = "host_setting"`. Present those as host-level setup guidance, not as edits to persistent Codex config.

Deep Security Scan uses MCP-owned SDK sessions rather than the parent thread's worker pool. Its preflight requires native V2 but does not derive outer worker concurrency from the parent session cap or require parent delegation depth. When migrating a legacy V1 configuration for Deep Security Scan, remediation removes `agents.max_threads`, sets `features.multi_agent_v2.enabled = true`, and sets `features.multi_agent_v2.max_concurrent_threads_per_session = 4`. The coordinator overrides each child session's cap from the per-user deep-scan configuration. Codex rejects the legacy V1 thread setting and explicit V2 mode together.
Deep Security Scan uses MCP-owned SDK sessions rather than the parent thread's worker pool. Its preflight does not require a particular parent delegation runtime, ownership, capacity, or depth. Discovery workers inherit the scan's model and run under the verified read-only worker sandbox.

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Retain the V1 compatibility gate for child sessions

When the active CODEX_HOME contains legacy agents.max_threads, this profile now returns ready, but the bundled CodexSdkWorkerExecutor.run still inherits that environment and explicitly passes features.multi_agent_v2.enabled=true to every discovery child. Codex rejects that V1/V2 combination, as the unchanged preflight documentation itself notes, so the newly admitted V1 configuration fails only after start_codex_security_deep_scan launches. Keep the config_absent check/remediation, or isolate the child configuration before declaring this parent setup compatible.

Useful? React with 👍 / 👎.

- `$codex-security:attack-path-analysis`

The discovery tool launches Codex workers that may use Subagents v2. The active configuration must satisfy the deep profile's native-v2 requirement. The worker count is configured separately from this thread's subagent allowance.
The discovery tool manages its own workers independently of this thread's delegation runtime and subagent allowance.

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Keep required phase skills in the deep preflight

Worker-pool independence does not remove the parent workflow's phase-skill dependencies: this same section still requires security-scan, threat-model, finding-discovery, validation, and attack-path-analysis, and the parent invokes them after discovery. In a runtime that exposes the Deep Scan entry skill but omits any of those selectively loaded skills, removing deep_scan_phase_skills lets preflight report ready; the expensive discovery can then finish before the mandatory centralized tail becomes impossible. Restore this skill-availability requirement independently of the delegation-runtime checks.

Useful? React with 👍 / 👎.

@chatgpt-codex-connector

Copy link
Copy Markdown

Security review completed. No security issues were found in this pull request.

Reviewed commit: 205fb5d2c1

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

  • You comment "@codex security review"
  • A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.

@mldangelo-oai

Copy link
Copy Markdown
Collaborator Author

@codex review

Please review the exact updated head c31d312baae2a83f4ad7ad0adbb7483c563b3ee5. This additive update ships the exact generated MCP runtime from the same original upstream commit and preserves the atomic base-to-head diff. The matching model-native worker executor now handles both inherited V1 and V2 configurations correctly.

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. Can't wait for the next one!

Reviewed commit: c31d312baa

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

@chatgpt-codex-connector

Copy link
Copy Markdown

Security review completed. No security issues were found in this pull request.

Reviewed commit: c31d312baa

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

  • You comment "@codex security review"
  • A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.

…/codex/model-native-deep-workers

# Conflicts:
#	sdk/typescript/_bundled_plugin/mcp/server.mjs.br.part-000
#	sdk/typescript/_bundled_plugin/mcp/server.mjs.br.part-001
@mldangelo-oai

Copy link
Copy Markdown
Collaborator Author

@codex review

Please review the exact updated head 6480979e93ecdfd01d96a95017bbbdafbb6e5476. Inherits the regression-tested prompt/headless rejoin fix and safe URL delegation; the generated MCP runtime matches this original upstream stage and retains the bounded headless user-context argument.

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. 🚀

Reviewed commit: 6480979e93

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

@chatgpt-codex-connector

Copy link
Copy Markdown

Security review completed. No security issues were found in this pull request.

Reviewed commit: 6480979e93

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

  • You comment "@codex security review"
  • A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.

…/codex/model-native-deep-workers

# Conflicts:
#	sdk/typescript/_bundled_plugin/mcp/server.mjs.br.part-000
#	sdk/typescript/_bundled_plugin/mcp/server.mjs.br.part-001
@mldangelo-oai

Copy link
Copy Markdown
Collaborator Author

@codex review

Please review the exact updated head 4b616a8c330c3d620a04f9ad49d14ae17696471f. Inherits regression-tested resumable finalization, accurate cached-token totals, distinct canonical finding identities, complete ignored-file coverage, safe context delegation, and preserved prompt/headless scan identity; the generated runtime remains matched to this original atomic upstream stage.

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. Hooray!

Reviewed commit: 4b616a8c33

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

@chatgpt-codex-connector

Copy link
Copy Markdown

Security review completed. No security issues were found in this pull request.

Reviewed commit: 4b616a8c33

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

  • You comment "@codex security review"
  • A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.

@mldangelo-oai

Copy link
Copy Markdown
Collaborator Author

@codex review

Please review the exact updated head 542a5a20b095269851e6b99113087d025172da25. Inherits regression-tested resumable finalization, accurate cached-token totals, distinct canonical finding identities, complete ignored-file coverage, safe context delegation, and preserved prompt/headless scan identity; the generated runtime remains matched to this original atomic upstream stage.

@chatgpt-codex-connector

Copy link
Copy Markdown

Codex Review: Didn't find any major issues. Swish!

Reviewed commit: 542a5a20b0

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

@chatgpt-codex-connector

Copy link
Copy Markdown

Security review completed. No security issues were found in this pull request.

Reviewed commit: 542a5a20b0

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

  • You comment "@codex security review"
  • A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant